What’s in a Name? Exploring Identity in the Field of Library Journal Publishing

INTRODUCTION This paper explores the variability in self-identifying practices of academic libraries engaged in journal publishing and hosting activities. We were interested in how libraries characterized their efforts in this area and looked at whether there is an unspoken threshold for differentiation with respect to publishing-support naming conventions. METHODS Using the Library Publishing Directory, in-depth interviews, and a more widely circulated follow-up survey, the research team examined service offerings, divisions of responsibility, funding, terminology, and semantic associations within publishing, both as an active practice and as an advertised service. RESULTS AND DISCUSSION We aimed to tease out whether there was any sort of tipping point, or inferred rules, around when an institution chose to call the activity either publishing or hosting. We found no particular service, set of services, funding structure, or division of labor that obviously influenced the use of a particular term. Rather than noting a divide between publishing and hosting, participants spoke of both a spectrum and a tiering of work and support, though all emphasized that these models did not describe the quality of the work produced. This paper also discusses how use of the term library publishing creates additional ambiguity in naming practices, and considers some implications for library staff newly immersed in scholarly publishing work.


INTRODUCTION
With the establishment of the Library Publishing Coalition (LPC) in 2013, academic libraries in North America were offered an opportunity to grow their community through organized support and leadership, and through the formal consolidation of efforts in this area under the umbrella term library publishing.
The LPC's Library Publishing Directory (Lippincott, 2015) offers an annual snapshot of the state of library publishing, as well as a permanent record of the growth of the movement. The Directory showcases a remarkable breadth of services, and the different combinations of scale and scope in publishing operations demonstrates the grassroots nature of many efforts in the field. Within the Directory listings, we noted a variance in how journal-focused services are named, and that a number of libraries offering seemingly similar services identify their offerings differently: where some describe their services as journal publishing, others apply the banner of journal hosting.
Our research team sought to explore the factors that influence library self-identification as journal publisher or journal host. Does the title publisher signify particular activities in the production chain of scholarly publishing? Or does the use of publisher as signifier reflect a particular level of institutional resource allocation and support?
This study explores how libraries engaged in this work characterize their efforts, looking at whether there is an unspoken threshold for differentiation with respect to publishing support naming conventions. Our team identified and tested for a number of factors that might contribute to a potential tipping point in naming practices, including institutional tensions and politics, a history of publishing at the institution, staff experience with publishing, staffing and service levels, and tolerance for risk or liability. While the term library publishing can be used to refer to a diversity of scholarly outputs, for the purpose of this study we have chosen to look specifically at journals, with the understanding that many institutions that produce journals also produce other formats.

LITERATURE REVIEW
Our research team was specifically interested in the distinctions made between publishing and hosting, and to this end has used the literature review to point to reports, case studies, and research where the relationship between those two terms, and implied underlying activities, is explored or explicitly characterized. There is ambiguity in the boundaries of what makes up "publishing services," but the Library Publishing Coalition characterizes its intended meaning broadly as "the set of activities led by college and university libraries to support the creation, dissemination, and curation of scholarly, creative, and/or educational works" (LPC, 2013, "About Us"). In both literature and conversation, the term library publishing has emerged as a moniker for work that is both a growing subfield of publishing and an expanding area of library responsibility.
While libraries have a long history of publishing some niche materials (see Canty, 2012;Okerson & Holzman, 2015), the development of the web and the shift to digital objects as the primary medium for exchanging research have created the conditions for libraries to take on a more central role in publishing operations. The last ten years have produced significant library literature around the growth of electronic publishing, with discussions typically tied to the open access movement, growth of web-based exchange of scholarly work, and increased software availability (e.g., Shearer, 2010;Hahn, 2008). In a landmark 2008 report for the Association of Research Libraries, Karla L. Hahn highlighted the fact that "research libraries [are] rapidly developing publishing services" (Hahn, 2008, p. 5). A 2010 survey by Morrison and Owen also noted such growth, with a wider focus that included non-research intensive institutions involved in journal publishing.
Academic libraries support publishing and hosting programs for myriad reasons, chief among them that the work is well aligned with their broader mandate, mission, or objectives. Park and Shim (2011), for example, demonstrate how library publishing satisfies Roosendall and Geurts's four functions of scholarly communication: registration, certification, awareness, and archiving (p. 77). Involvement in publishing also affords greater chances for input into licensing and potential intellectual property constraints, giving libraries more say in the man-agement and preservation of digital scholarly publications (Skinner, Lippincott, Speer, & Walters, 2014).
However, what exactly constitutes publishing is not something that there is a lot of agreement on. In an article on the development of the Library Publishing Coalition, Sarah K. Lippincott (2016) notes that there is "difficulty of establishing clear boundaries around what constitutes 'library publishing'" (para 6). This is not a criticism of the work or those involved, but reflects the way programs have "emerged organically over the past several decades," as libraries "mindfully and strategically [shift] their roles on campus" (para 3). Lippincott concludes that "Library-based publishing is perhaps better defined by its motivations than its outputs" (para 5). This suggestion points to two questions libraries must wrestle with concerning the nature of publishing-a question of inputs (what labor makes us a publisher?), as well as one of outputs (what end products count as publications?). That library publishing is a newer subfield and so often explicitly mission driven arguably gives the scope of the work greater flexibility, allowing it to expand and contract around both of these questions.
In The Once and Future Publishing Library (2015), Okerson and Holzman point out that even those at the center of the library publishing world are not in agreement: "The boundary between activities that merit the name publishing and less formal and coherent enterprises is fluid and contestable. Discrepancies are evident even between the LPC's definition of library publishing and the activities listed in the directories by members" (p. 9). In discussing publishing as a spectrum, Okerson and Holzman argue that the large and well-funded end of the spectrum is easy to define, but "there is no agreement where to draw the line for the other end of the spectrum" (p. 9), pointing to grey literature, data sets, and institutional repositories as examples of what might or might not be publishing. Okerson and Holzman are less concerned with labeling publishing than in being able to identify "significant areas of overlap and areas unique to certain parties," a task made difficult when many types and levels of effort are characterized using the same terminology (p. 9).
There is a dearth of literature that focuses specifically on the differentiation of publishing and hosting work, but the relationship between the two practices is discussed in research. In her ARL report, Hahn (2008) characterizes library publishing as significantly scaled back from traditional publishers, arguing that "library publishing services have few pretensions to the production of elaborate publications" (p. 6). She instead characterizes the type of work being doing at most libraries as "light overhead publishing -based on open source software, close integration with related services, maximum use of digital formats, and a minimum of bells and whistles" (p. 24). More explicitly, she states: "'publishing services' may immediately conjure hosting services-server support, software development or installation," suggesting that hosting services are the core of any publishing services (p. 15). In this instance, there is no distinction made between hosting and publishing, because hosting is the essence of digital publishing.
Conversely, Morrison and Owen (2010) discuss "journal hosting" as one of many activities libraries can do to support open access publishing, such as budgeting for APCs or hosting preprints in an institutional repository (p. 32). They too discuss a "spectrum" of support, but here the spectrum is one of levels of hosting, with "full blown publishing support with an extensive suite of editorial services" being the far end of the hosting spectrum (23).
In her 2010 report "A Review of Emerging Models in Canadian Academic Publishing," Kathleen Shearer characterizes library services as "typically involv[ing] basic activities like hosting, technical assistance, and training in the use of the journal publishing software" (p. 11). She does note that one school is "offering an extra level of support beyond the common hosting services being offered by other libraries" (p. 14), but her other library case studies focus on software development or technical infrastructure support for journal publishing. Another seminal report written by three library deans, Library Publishing Services: Strategies for Success (Mullins et al., 2012), examines the rise of publishing as "a core activity of North American academic libraries" (abstract). The paper includes reports from a number of workshops, where participants were characterized as often running "skeletal production systems and minimal editorial support," but are finding demand for much more in-depth support (p. 15). Indeed, the report authors specifically encourage expansions of partnerships in order to "meet requirements beyond mere software and content hosting and barebones staffing that have characterized some of our earliest forays into providing publishing services" (p. 3).
Here there is a recognition that hosting does not meet the needs of the user community, and the authors emphasize a need for training in areas such as XML workflow development and project management, including via library schools (p. 15).
In some cases, a continuum between publishing and hosting is made explicit. In What to Expect When You are Not Expecting to Be a Publisher (2014), De Groote and Case describe their journey from adamantly being "a host, an electronic distributor" (p. 168) to finally "admitting" to being a publisher (p. 173). Their initial scope of "providing a cost-effective platform for existing journals" (p. 168) was slowly transformed as policies and MOUs developed, and faculty began to make requests that "challenged our thinking on whether the library was hosting journals or publishing journals" (p. 170), which was further pushed by MOUs with indexing and registration services that required that a publisher be named. (p. 173). That paperwork pushes libraries to consider their role as publisher explicitly also came up in our own research.
We see from the literature that the terminology around library publishing is fluid, and matches the evolving nature of these services within libraries themselves. The diversity of these interpretations and distinctions informed the structuring of our exploratory research.

METHODOLOGY
We adopted a mixed-methods approach to provide both qualitative and quantitative data in our exploration of library identity around publishing support practices. Research instruments were submitted for ethics review at our respective institutions. The sharing of interview quotes is provided anonymously, and with the permission of interviewees.
Our first step was to develop a sampling frame, which consisted of 124 academic libraries in the United States and Canada that self-identified as supporting journal publishing or hosting activities. Data for the sampling frame was drawn from the 2014 Library Publishing Directory, from past Synergies 1 member institutions, and from Ontario Council of University Libraries institutions. We noted number of titles published, formats published in, platforms used, and staff makeup.
An interview instrument was drafted based on preliminary analysis of the sampling frame. The interview was piloted with three institutions, and revised based on participant feedback and research team observations. The updated instrument was then used to interview eight individuals involved in library publishing or hosting (see Appendix A). Our semistructured interviews were geared to explore the breadth of opinion pertaining to activities supporting journal publishing in libraries. To that end, our sampling strategy was not random; it aimed to include larger and smaller institutions from both the United States and Canada. Half our sample were libraries that self-identified as host in some obvious way, with the balance identifying as publisher. We also targeted institutions with mature programs publishing numerous journals. Interviews were recorded, transcribed, and submitted to interviewees. Interviewees were given the opportunity to read their transcripts and provide corrections. Corrected interviews were each coded twice with Dedoose research analysis software to extract qualitative findings pertaining to key themes.
After interviews were completed and analyzed, we constructed a questionnaire to test for the generalizability of key findings. We expanded our sampling frame against the 2015 edition of the Library Publishing Directory and the questionnaire was distributed by email. We contacted 144 institutions, and the survey generated a 47% response rate. The survey instrument is included in Appendix B.

RESULTS
Our research produced several pools of data: a sampling frame, interview transcripts, and survey responses. In developing the sampling frame, we found that it was sometimes difficult to discern how institutions self-identified, and we needed to contact 20% of institutions to clarify whether they identified as a journal publisher or host. We were unable to confirm a term for 3% of the sample.
Figure 1 below shows the breakdown of self-identification in our sampling frame. Slightly over half of our sample identified as host, 34% identified as publisher, and 8% called themselves both. Interview participants and potential survey respondents were drawn from this sampling frame. For clarity, in the following discussion we have distinguished between "interviewees," those who took part in our semistructured interviews (eight people), and "respondents," those who responded to the survey (67 people total). We refer to the two groups collectively as participants.
We began both the interview and the survey with some questions designed to contextualize responses. Most interviewees indicated that their institution's publishing program had been in operation for between 6 and 10 years, while their own experience with publishing ranged from decades to only a few months. Similarly, in our survey 64% of respondents indicated that their service operations had been in existence for over 6 years, while 15% had been running for less than 2 years. Respondents echoed what was observed in our sampling frame with respect to staffing, with 63% having small teams focused on publishing efforts, 20% having a large team dedicated to this work, and 17% being the sole employee of the service. Half of respondents reported a growth in resources devoted to journal publishing since their program had begun; 8% of respondents reported having fewer resources than when the program began. The amount of resources dedicated to the programs had no significant effect on whether libraries identified themselves as publishers or hosts.
We wanted to explore with participants the meaning they associated with the terminology used in this area of work. We specifically asked our interviewees whether they saw a difference between the terms publishing, library publishing, and hosting. Some interviewees indicated they were uncomfortable with the term library publishing because they did not see a case for including library as a qualifier. One interviewee plainly stated, "to me library publishing is not its own category" (Interview 5). Another interviewee explained, "If you want to compete, you can't say 'well we're just a library publisher.' No you're a publisher-you're a publisher as good as any other and that's what we want to be for our clients." (Interview 4). Another interviewee wondered if the terminology invited scorn: "People are just calling it 'Library Publishing' and at least people who work for presses are kind of . . . I would imagine that maybe some of them look down on library publishers as not being real publishers" (Interview 7).
Regardless of terminology used, there was an emergent theme in interviews that library goals and values were a differentiating factor. One interviewee stated, "I don't believe in the distinction between library publisher and publisher. Now that doesn't mean I think that we have nothing special to offer as a library . . . because of our ethos as a library." (Interview 4). In distinguishing between "journals that are just plopped on our platform" and "library publishing," another interviewee stated: "The biggest difference is the educational component . . . and also the contribution back to departments and to the institution to support the original research that's going on" (Interview 3). Other interviewees identified their focus on open access as a key pillar of their work: "Quality is a major issue in fighting the misconception that faculty have around Open Access journals . . . . This is another reason to be intentional in what you are publishing and how you support it" (Interview 2). In regards to how they self-identified, our interviewees were split down the middle, with half seeing their role as publishers quite clear-cut, with responses such as "we are definitely a publisher" (Interview 1) or "I've been getting catalog records corrected and making sure the library is listed as publisher" (Interview 2). Our other four interviewees had less definitive responses, such as "We don't consider ourselves, right now, a publisher per se" (Interview 7), or "What we're doing can be described as hosting . . . you could go either way on that one to be honest" (Interview 6), or "We are really giving them technology support and moral support and helping them think through the issues where they don't necessarily have expertise. . . . I wouldn't call that publishing" (Interview 8).
Our follow-up survey asked similar questions, though often simplified to allow for yes/ no responses. A third of survey respondents indicated that they did not see a meaningful distinction between the terms publisher and library publisher (see Figure 2), while 40% of respondents indicated the distinction made sense only with particular stakeholder groups. Although we did not include anything about "hosting" in these questions, two respondents wrote in comments saying they preferred to be called hosts, and another written-in comment hinted in the same direction: "I wish there was another term for library publishing since we don't do everything a publisher does." The idea of a spectrum or continuum of services was introduced by interviewees again and again: "If we're thinking about it as a continuum, we're probably sort of in between hosting and library publishing" (Interview 3). Some interviewees and survey respondents characterized it as a publishing-hosting spectrum ("I really think that the services we provide are essentially between hosting and publishing-enhanced hosting or something" [Interview 7]), while others talked of a publishing spectrum that was separate from basic hosting services ("I don't feel like publishing has to mean one thing, although I do feel like it's distinct from hosting" [Interview 5]).
While the "Capital P" publishing end of the spectrum was described in fairly similar terms in all mentions by interviewees, it was the other end of the spectrum that was less clear, particularly with respect to where hosting might fit, as well as broader questions of what counted as publishing ("I also find it interesting to think about digital humanities projects and how those fit in" [Interview 7]) or "Is Wordpress.com a publisher or are they a hosting platform?" [Interview 8]). This lack of clarity around the relationship between hosting services and the word publishing echoes the analysis of Okerson and Holzman in The Once and Future Publishing Library (2015), as discussed in the literature review.
Following the exploration of terminology with participants, we turned our focus to the relationship between service offerings and the designate of publisher or host. Interviewees were shown a chart that detailed the frequency with which services were offered across institutions listed in the Library Publishing Directory. We asked them if they saw services they didn't offer but would if they had larger budgets, and conversely if there were any listed services they would never be interested in offering. There was wide divergence in these answers; the most notable split in opinion was around software development, peerreview management, and marketing. These services were identified as being of high interest to several interviewees, and of little to no interest to several others. Where software was mentioned as an area of future interest, two interviewees wanted to focus on expanding the development of software currently in use, while two others favored a focus on new forms for research outputs. One interviewee explained that they were interested in "more adaptive and expansive systems to provide platforms for new modes of scholarship" (Interview 2). Another interviewee emphasized the unique work libraries were well-positioned to do: "In terms of building up new things . . . I think there's a bigger role for libraries to be involved doing that and helping scholars to think through how do you build something that's not a linear text?" (Interview 8).
In our follow-up survey, we asked respondents whether they considered particular services to be characteristic of the designate library publisher or library host, with the option to select one, both, or neither for each service (see Appendix B for list of characteristics). Our aim was to look for a tipping point between these two concepts. The most dramatic divergence between the two concepts was with the tasks of copyediting, quality control of the scholarly content, and marketing: 100% of survey respondents said these activities were characteristic of publishers, while very few thought they were tasks a host would undertake. But for the libraries in our sampling frame that explicitly call themselves publishers on their website, only 32% offered copyediting services, and 62% offered marketing support. This suggests a possible disconnect between expectations or ideas of what a publisher does and the actual scope of work performed. This finding casts doubt on one of our initial hypotheses, which tied library service naming practices to services offered. It also seems to support a comment from one interviewee that the word publisher was more of a signifier than a job description: "[If ] you actually have substantive contributions to making decisions about how and when and where things get published, you're probably acting as a publisher" (Interview 1).
Similarly, 98% of survey respondents indicated that managing the journal's design was an activity characteristic of publishers, while in multiple interviews, graphic design was explicitly identified as not being of interest, regardless of budget. Other categories of services, such as the vetting of participating journals, registering journals with abstracting and indexing services, and DOI registration were seen as characteristic of publishers by most respondents (92% or more), while around half the respondents saw these as part of hosting work. User training (83% publisher / 81% host), the provision of statistics (86% publisher / 85% host), and preservation (82% publisher / 83% host) were services seen as equally characteristic of publishers and hosts.
While some interviewees and survey respondents pushed back at the suggestion that publishers can be defined based on a particular suite of services, there was broad acknowledgment that tiers exist in the publishing continuum, but just as importantly that tiers were not indicative of the value of work produced. As one interviewee put it: "I see the temptation to say 'Ok there's a spectrum-where can we put markers to sort of indicate the levels of what we're doing?' But I hesitate to do it because I'm afraid then that tier 5 with super services would then come to express that that's where all the quality is, and I don't think that's true" (Interview 5). Another linked tiering to research disciplines: "Different disciplines have innate expectations of what a tier looks like and what a publisher should do for them" (Interview 1). This was echoed in survey respondents' answers to a similar question, with comments noting that "library publishing is all over the spectrum" and that some sites offer services very similar to commercial publishers.
To see whether codifying of relationships affected the way participants perceived their service level, we asked participants how they described themselves in formal agreements with their journals or with third-party services. While four interviewees definitively self-identi-fied as publishers, only two of them were insistent that they be named as the publisher of record. In most other cases there was a looseness around this association. One interviewee stated that their institution preferred to be named publisher, but did not require it: "We feel that it's fair to say that we're the publisher. But we have kind of backed off on that if there are cases where it bothers editors." (Interview 5).
Many of the interviewees stated that the process of solidifying the relationship between journals and the service unit was underway. In several cases there was interest in formalizing the relationship: "No we don't actually have any current agreements. But in the future I can say unequivocally that the service level agreements and the MOU we use will most definitely refer to the library as a publisher and beyond that we're going to demand, for lack of a better word, that journals specifically reference us as their publisher" (Interview 1). In another case, the transition was toward a more flexible approach. One interviewee at a large and well-established institution described their new and less-rigid MOU as follows: "Our agreements used to be very formal, legal documents. One of the reasons we've simplified the agreements is that the authors and editors we're working with are often totally on their own. Our policies seek to be friendly and helpful to them-so an agreement full of intimidating language doesn't really serve that end-especially if it's not actually necessary for journals that are also based at our university. We've moved instead to a simpler documentation of our shared responsibilities" (Interview 5). Two interviewees noted that binding departmental journals with a legal agreement didn't make sense, as they were all one legal body. One likened trying to enforce such a legal agreement to "suing yourself." Instead, these interviewees spoke of using their documentation to clarify the roles and responsibilities of each of the parties, laying out how journals and libraries could expect to work with each other.
Our survey results also point to fluidity in terminology with regard to asserting themselves as 'publisher' in formal agreements with journals. Of all survey respondents, 60% reported having a written agreement with their journals, and of those only 50% identified their unit or library as publisher in these documents. When supplying metadata to an external service, only 12% of survey respondents insisted they be named as publisher, 28% said they would always name the journal, and 22% said they would defer to a journal's preference. In survey comments, we heard responses that echoed the mixed feelings of our interviewees in being named as publisher, for example: "We would like to avoid legal/administrative consequence of the library being officially listed as 'publisher,"' and "The question about whether we ARE the publisher for our hosted journals is still not settled." We also wanted to explore whether campus cultures, institutional tensions and politics, or other on-campus services shaped both the scope of work and the naming practices of publishing or hosting units. Interviewees were asked whether there were political or admin-istrative pressures that determined what services they could offer within their unit. Most reported that their institution was fairly hands-off in determining what could and could not be offered. In two cases, university counsel vetted the service scope articulated in MOUs. At one institution, competition with on-campus editing services prevented the provisioning of a copyediting service by the journal publishing program.
When asked about library and university comfort with taking on the responsibility for publishing scholarly content, half of those interviewed identified no conflicts in the area. Of the others, one institution anticipates potential pushback from the dean in the event that the library has an interest in being named publisher. Nonprofit status factored into some of the answers; in two cases, the inability of the library to accept money affected the scope of their work. One interviewee reported concerns from the university over the quality of peer review reflecting on the broader institution if it were officially named publisher.
The ways in which our survey respondents described their services to stakeholders, whether to on-site scholars, the broader university community, or the library community, were wide ranging. Respondents were able to choose more than one term, which prevented us from being able to filter by a single dominant affiliation and perform a cross-comparison, though this confusion clearly best represents the fluidity in both terminology and scope. Almost 30% of respondents described themselves as providing hosting services only, 13.5% described their services as publishing only, and 13.5% described themselves as library publishers only; 32% used some combination of these three terms to describe their services. The 10% of respondents who used none of these terms identified their services with terms such as digital scholarship services, digital collections, open access (journal) publishing or scholarly communication(s), and publishing.
Finally, we asked participants about the genesis of the formal name of their unit, and whether it had been a conscious exercise. Findings from our interviews were split between deliberate exercises in name generation and name inheritance by virtue of software or because a name was offered under a particular umbrella of services. In the cases where naming was a deliberate practice, a number of reasons were put forward: one institution pointedly wanted to include "publisher" in the name, and another wanted to distinguish its offerings from those of their university press. Other reasons for intentional naming practices included interest in establishing a reputation and the desire to be clearly descriptive of service offerings. Our survey asked respondents to select influencing factors on their name choice. Of all respondents, 16% identified administrative or political elements as a determining factor in their choice of name for their service. Similarly, legal considerations were not highly reported as a determining factor at only 5%, nor was funding at 13%. At 22%, alignment with university or library mission was a slightly more important factor affecting naming practices. The factors that most often affected naming choice proved to be the most obvious: the number and nature of services offered (45%) and the fact that the name best described the service (72%).
One third of survey respondents reported having had put "some thought" or "a lot of time and thought" into naming their journal publishing service (Figure 3a), while 25% of respondents said they didn't have a single distinct name or identity. This correlates with what we found when building our sampling frame-we needed to reach out to 20% of institutions to clarify the name of their unit, and whether they self-identified as a publisher. Figure 3b illustrates a breakdown of the use of several key terms by the survey respondents. Overall, the name of the university was incorporated into the service name almost half of the time, the word library was present in a quarter of the sample, and the word publisher or publishing appeared in the names of a little over a third of the sample. When results were filtered to analyze institutions that reported being more deliberate with their naming practice (those that had put "some thought" or "a lot of thought" into their name), we saw a tendency toward developing a brand name that did not include the word library: the use of the word library fell from 26% to 5%, and the adoption of a brand name rose from 30% to 50%.

DISCUSSION
Our research was structured to surface a plurality of opinions, and this was indeed the result-across all data collections we could find no obvious or even suggested tipping point, line, or niche that separated publishing and hosting. That said, we did see some common themes and areas of agreement emerge. First, there was consensus from journal publishing initiatives both large and small that publishing is a spectrum with flexible boundaries. Second, the level of formality with which libraries take on the work of supporting publishing varies quite widely between institutions, regardless of their years of experience in the area. And finally, we found that the ethos and mission of the library does inform the breadth, scope, and types of support offered.
Our findings point to a few possible takeaways for library staff newly engaging in publishing. For those at institutions that are just beginning this kind of work, it would be useful to determine the level of formality one might wish to articulate in a service agreement from the outset. In both interviews and survey responses, we heard how challenging it was to try to introduce formal policies or MOUs once a journal was in production. This often resulted in a pastiche of paperwork, where newer journals were subject to different policies than those that operated "the way it's always been." Of course, there were cases where this flexibility was highly valued; not all journals need the same amount of support, and offering a custom experience for each journal may simply be part of the local service ethos.
More obvious implications arise for individuals joining institutions that have publishing programs in place. Establishing guiding principles around how the unit wants to self-identify when courting faculty, applying for funding, and engaging with student groups could lend clarity to expectations and understanding of scope of support. Many interviewees indicated they didn't have any documented scope-they just knew what they could and couldn't do. For those who are new to an institution, an understanding that the scope is often an intangible cultural factor that needs to be learned informally may be reassuring. It might also be a nudge to ask more questions.
An issue that was surfaced and deserves more exploration is the use of the term library publishing. When we began this project, our focus was centered on comparing and contrasting publishing and hosting. In our pilot interviews it became clear that we needed to make space for conversations outside of this binary, because the concept of library publishing kept being introduced. We therefore included library publishing as a separate category in both our interview and survey instruments. We found a fairly clear-cut divide between those who saw the phrase library publishing as a meaningful way to distinguish their work and those who saw it as loaded or unhelpful. Participants certainly agreed that a publishing continuum exists, but there is a blurring in how the publishing work in libraries is described on this continuum, and we suggest that this is likely exacerbated by a complex relationship with the term library publishing.
Along with parsing the categories, we were challenged by the linguistic gymnastics required to abstain from naming an activity as publishing or hosting in our own descriptions. Where we discussed the spectrum, publishing often proved the easiest term to use, thus importing a certain bias to our conversations and writing.
Another limitation of this research is the time lag from initial sampling frame development to publication, given that library publishing is a fast-changing field. Our initial sampling frame was drawn up in 2015, based on 2014 data. It is not difficult to imagine some fairly significant changes in service offerings having taken place between then and now. An analysis of the Library Publishing Directory's service census over time could bring to light the way that this transformation is taking place.
The question of whether an institution identifies as a publisher or a host, both, or something in between is bound up in larger questions of the role of the library in the production of scholarship. Our aim was to explore the shifting boundaries of the field, to report on some of the deliberations that have led us to where we are, and to contemplate how these conversations and practices impact how we think about ourselves. It matters what we call ourselves; the very fact that so many respondents changed their naming convention for different audiences tells us so. The lack of cohesion in service offerings and ambiguity in naming practices is not, of course, indicative of any crisis of identity. A sameness of service types across institutions will never make sense, and less so for emerging services.

CONCLUSION
This study aimed to examine the factors influencing naming practices of library journal publishing programs, and to uncover the factors influencing choices of names or labels. As the study progressed, the research team found the nuance and depth of meaning in this area of inquiry was greater than initially anticipated.
We set out to explore the factors that influence library self-identification as either journal publisher or journal host and found that service naming was not related to any set of particular activities being conducted. Instead, we found that while hosting and publishing are often discussed as existing on opposite poles, the actual labor associated with each term is much more flexible and interconnected. The process of constructing a sampling frame gave us our first clue as to the complexity of this ecosystem, where a full 20% of institutions did not clearly call themselves either a publisher or a host on their journal publishing-related information pages. An outcome of our pilot interviews was the introduction of a third potential category, library publishing, which invoked very mixed reactions from participants.
Our attempts to pin down criteria for self-identification as publisher or host during interviews did not bring further clarification. Initially, we thought this was due to our purposeful selection of diverse interviewees, as we were hoping to surface varied voices to inform the synthesis of our survey instrument. The results of our survey, built to account for and test for the prevalence of these additional viewpoints, confirmed that our research question had, instead of being quantifiable and resolvable, unearthed a number of further questions, and a rich field for exploration and discovery.
In addition to giving us a temporal snapshot of practices and philosophies around library publishing, our study proved to be an exercise in cataloguing the different ways that libraries have organized themselves to serve the unique needs of their communities, with many different levels of resources and expertise. The question of how to self-identify is being wrestled with by organizations who have been publishing for many years, as well as those just starting their first journal, by those with large staff complements, as well as those with single employees dedicated to their programs. As was made clear through our research,

APPENDIX A Semistructured Interview Instrument
(1) What is your role at the institution, and how much of your time is dedicated to scholarly publishing? (Prompt for details.) • How long have you had scholarly publishing responsibilities, and has your workload always been split like this [or 'focused there' if all they do is scholarly publishing]?
(2) Do you have any past experience with publishing?
• Was there a learning curve for you with your current position?
• How did you catch up? (Prompt for details like specialized training, etc.) (3) Do you see a difference between publishing, library publishing and hosting?
• Do you believe there to be a minimum subset of services a publisher should offer?
• Does a publisher need to be full service?
• Are there different tiers?
(4) Have a look at the spreadsheet we sent you. It gathers data from the Library Publishing Coalition to show the percentages of libraries offering different kinds of services. We've grouped them by function for ease of reading. Are you surprised by the frequency of any of these offerings? Why?
• When you talk about your own services (to on-site scholars, to the broader university community, to the library community), do you make it clear which of these services you do not offer?
• If you had unlimited resources, Which of these services would you like to offer?
Which would you not want to offer?
(5) Were there any political or administrative pressures that affected what you could offer? • Did the university or library not want to take on responsibility for publishing scholarly content?
• Were there limitations (financial, staffing, liability) to what could be promised in an agreement?
• Is there potential competition with a university press?