Scaling Small; Or How to Envision New Relationalities for Knowledge Production

Within the field of open access (OA) publishing, community-led publishing projects are experimenting increasingly with new forms of collaboration and organisation. They do so by focusing on setting up horizontal alliances between independent projects within a certain sector (e.g., scholar-led presses), or vertically across sectors with other not-for-profit organisations (e.g., through collaborations with libraries, universities, and funders), in order to create multi-stakeholder ecologies within scholarly publishing. Yet at the same time, imaginaries for future modes of OA knowledge production are still controlled through demands for ‘scalability’ and ‘sustainability’, which are both seen as preconditions for scholarly communication models and practices to succeed and to be efficient. But they are also prerequi­sites to receive funding for publishing projects or infrastructure development. The scalability of open models is perceived as essential to compete in a landscape dominated by a handful of major corporate players.Drawing on our work with the Radical Open Access Collective, the ScholarLed consortium, and the Community-led Open Publishing Infrastructures for Mono­graphs (COPIM) project, this article outlines an alternative organisational prin­ciple for governing community-led publishing projects based on mutual reliance, care, and other forms of commoning. Termed ‘scaling small’, this principle eschews standard approaches to organisational growth that tend to flatten community diversity through economies of scale. Instead, it puts forward the idea that scale can be nurtured through intentional collaborations between community-driven pro­jects that promote a bibliodiverse ecosystem while providing resilience through resource sharing and other kinds of collaboration. Following Anna Tsing’s recom­mendations to keep in mind how reimagining our knowledge practices requires we pay particular attention to articulations between the scalable and the nonscalable (Tsing, 2012), what is needed to enable this is, first and foremost, a rethinking of existing systems and infrastructures and how they currently function – systems that have historically developed and been continuously remade to encourage fur­ther scalability. We further explore the possibilities of scaling small with particular reference to Anna Tsing’s work on the ‘latent commons’ and Massimo De Angelis’ discussion of ‘boundary commoning’, examining how these concepts are on display within the Radical Open Access Collective, ScholarLed and the COPIM project. As we will argue, reimagining the relations within publishing beyond a mere calcula­tive logic, i.e., one that is focused on assessing the sustainability of alternative models, is essential in not-for-profit OA publishing environments, particularly if we want new forms of collaboration to arise and to redefine the future of scholarly publishing in communal settings.

our experiences of scaling small within different publishing collectives, we reveal the different ways in which small OA presses can become stronger through mutual reliance and other kinds of collaboration.
In deploying the language of 'commoning' and the 'the commons', we hope to emphasise the value of relationalities within projects that collaborate on and manage shared resources. We borrow from Massimo De Angelis and Stavros Stavrides who position the commons as a mode of production that prioritises the self-organisation of labour over market and state forms of production (De Angelis and Stavrides, 2010). Taken in this way, the commons concerns not just the resource and the rules for its governance, but also the ways in which commoners relate to one another and the subjectivities nurtured in doing so (Moore, 2018). Drawing on Anna Tsing's notion of the 'latent commons', we discuss how these practices of commoning are present but underdeveloped within the ROAC. We then imagine what more formalised collaborations between OA book publishers might look like based on Massimo De Angelis' concept of 'boundary commoning', a theory that explores how projects can cultivate mutual reliance by embedding themselves in one another's projects. In doing this, we therefore hope to complicate those understandings of the commons that focus primarily on resources and the rules for governing these resources, arguing instead that relationalities and cultivated subjectivities are just as important to appreciate within shared modes of production.

Sustainability and resilience
For more than a decade now, experiments in OA book publishing have been plagued by the demand for a 'sustainable and scalable' alternative to print-based subscription models, where scale and sustainability are upheld as essential requirements for a successful OA business or publishing model. Professor Margot Finn, President of the Royal Historical Society, talks about the need to overcome obstacles in order to have ' a sustainable, scalable, global open access model' (Finn, 2019). The librarian Rick Anderson argues that to establish 'proof of concept' for ongoing publishing projects or initiatives, 'proof of program' (sustainability) and 'proof of scale' are essential (Anderson, 2015). Connected to these demands for scale is the underlying presumption that only large publishers can scale or benefit from economies of scale. For example, the publishing consultant Michael Clarke argues that: 'A small society publishing program cannot muster the economies of scale around production, technology, and (most important) institutional sales that can be brought to bear by a large publisher' (Clarke, 2015). Meanwhile, the publisher Richard Fisher states that, ' at present, we have numerous, very welcome small-scale Open Access publisher experiments, but these remain artisanal, cottagesized responses to what is a global challenge of academic production (overproduction) on a massive industrial scale' (Fisher, 2015).
But what does sustainability or scalability actually mean within the context of OA book publishing? How has it been framed within this specific context? For example, does sustainability look to a single (one to rule them all) economic model being sustainable, or does it require making the entire system of scholarly publishing sustainable? Does it mean commercially viable, making a profit, or breaking even? Is it about being sustainable in the short or long term? And sustainable for who exactly? For scholars and their institutions? For governments and taxpayers? For stakeholders in the commercial publishing model as it is currently set up? Does it mean self-sustainable or sustainable with a certain amount of (public) funding? And does sustainability mean the same thing for journals as it does for books, and for the sciences as for the humanities? Are we talking about sustainability in the sense of our relationship as publishers and scholars to technology, the environment, or the wider ecology of which we are a part?
The search among publishers, funders, governments, and other stakeholders within scholarly communication for the sustainable and scalable OA business model has led to a situation where paying fees for OA article publishing is becoming the norm. Similarly, increasing numbers of both publishers and funders are inclined to use or promote BPCs (book processing charges) for OA books, too (Penier, Eve and Grady 2020;Speicher et al., 2018), where the recent rapid uptake of BPCs by commercial book publishers seems to convey the message that this is the quickest, most sustainable, least disruptive, and most accountable way to reach universal open accessibility for monographs. Yet does this not provide a form of 'sustainability' that perhaps predominantly serves publishers and maintains the publishing system as it is currently set up? Do we need to take into consideration that, from the point of view of many incumbents, what actually needs sustaining are large profit margins (Ottina, 2013: 608)? What about authors and their institutions? A reliance on publication fees (which are generally covered by institutions or projects on behalf of the author(s)) risks disenfranchising independent, non-affiliated or so-called para-academics, early career researchers and PhD students, and those on casualised contracts; but also scholars from the so-called 'Global South' or working in less wealthy institutions, or those who create the kind of research that critiques the institutions of which they are a part, expressing viewpoints that do not meet with institutional approval. 2 Scholar-led publishers tend to explore a wide variety of different options to cover publishing costs, adapting to changing situations and new collaboration and funding opportunities, taking in everything from print sales to crowd-funding, from consortial funding models to donations, from volunteer and gift labour to institutional support and one-off grants (Adema and Stone, 2017;Gatti, 2015;Hall, 2015;Oei, Joy and Rudmann, 2020). When it comes to book publishing, and OA book publishing in particular, this is not a situation that is unique to scholar-led presses. Many, if not most, publishers of books in the Humanities and Social Sciences (HSS) work with these kinds of mixed models and cross-subsidies (instead of a single business model): both to make ends meet, and to publish the kinds of works that have always been hard to market let alone make profitable, whether it be due to their specialised focus, niche topic, experimental multimodal outlook, or their critical or radical character (Penier et al., 2020: 5;Speicher et al., 2018: 6). For books in the HSS, sustainable business models are in many ways an oxymoron. As the recent COPIM report on revenue models for OA books states: 'it is notable that a great deal of scholarly monograph publishing is subsidised, which should cause us to question the potentially problematic rhetoric of "sustainability"' (Penier et al., 2020: 5).
This is why people such as Open Humanities Press' David Ottina have argued that 'sustainable publishing' invokes the wrong frame for understanding the stakes and dynamics of scholarly communications, which is actually much more of a volatile endeavour than a stable one. Especially with respect to the publishing of monographs in the humanities, one could argue that it has never been economically self-sustaining, as it has always relied on some form of additional external funding and support. A frame of resilient communications might fit these dynamics better. The OA movement has already put forward many independent projects, which, if we take them all together, are developing what Ottina calls a robust and resilient system of scholarly communications. It is exactly the diversity of these initiatives, next to being small and relatively flexible endeavours that are not risk averse, which enables their capacity for change and experimentation. For Ottina, instead of focusing on economic sustainability, we should instead focus on a discussion of 'what kind of scholarly communications system we want and how we can make it resilient in the face of technological, institutional, and funding volatility' (Ottina, 2013: 209). Ottina provides the example of Open Humanities Press as an example of 'small-scale resilience'. Accordingly, we will argue that -building on Ottina's ideas -'scaling small', or the move towards creating (loose) alliances, collectives, and cooperatives between scholar-led and not-for-profit publishing entities, similarly stimulates the creation of these more resilient structures. To further outline what this entails for us and how we see this applied in the publishing projects that we support, we first want to draw further connections with Anna Tsing's writing on scale.

Scalability and nonscalability
The anthropologist Anna Tsing has developed a valuable theorisation around scalability that focuses on the associations between the scalable and the nonscalable. One of the main points of her argument is that scaling or scalability assumes that projects expand smoothly without changing their framing assumptions (Tsing, 2017). For example, within a business context this refers to the ability of a firm to expand without changing the nature of what it does or without changing its organisation. Economies of scale would comprise one such kind of business scalability (Tsing, 2012). The aim here is to make projects more scalable for expansion -for growth, profit, and progress -and the desirability for doing so is taken for granted, Tsing argues. In other words, to scale up or to scale well implies developing the quality called scalability, which is the ongoing ability to expand without rethinking basic elements. Yet it is important to emphasise that the intention here is not only to improve but also to naturalise scalability. According to Tsing this illustrates how the discourse of scaling up has been strongly connected to our ideas of progress and to the advancement of humanity, where the success of expansion through scalability has shaped both modernity and capitalism -and, with them, knowledge production. From this perspective a system that is sustainable ought to be scalable, and nonscalable systems become an impediment to progress.
The issue with scalable projects, as Tsing points out, is that they tend to be oblivious, both to the diversity of contexts, and to the indeterminacies that originate from encounters with this (cultural and, in the case of her book on the possibility of life in capitalist ruins, biological) diversity. Scalability allows us to see only uniform blocks, ready for further expansion, blocking our ability to notice the heterogeneity of the world and disguising its divisions. As Tsing explains: Scalability is possible only if project elements do not form transformative relationships that might change the project as elements are added. But transformative relationships are the medium for the emergence of diversity. Scalability projects banish meaningful diversity, which is to say, diversity that might change things. (Tsing, 2012: 507) The urge for scalability has transformed both our societies and landscapes -and similarly our knowledge systems and infrastructures -but it hasn't been able to suppress nonscalable economic and ecological relationships, which continue to erupt in unexpected places (Tsing, 2017: 42). Yet Tsing does not see the nonscalable as existing in a normative relationship to scalability here (i.e., as one being good and the other being bad). The main distinguishing feature between scalable and nonscalable projects is not ethical conduct; it is rather that the latter are more diverse because they are not geared up for expansion. Scalability banishes meaningful diversity, that is, diversity that might change things.
Therefore, to provide an alternative way to conceptualise the world beyond expansion, Tsing argues for the development of a 'theory of nonscalability' (Tsing, 2012: 505), which shows that scalability uses and depends on nonscalable forms, even if it denies or erases them. Within scholarly publishing we see this, for example, in the staggering amount of volunteer, free, or underpaid academic labour that commercial publishers depend on as part of their profit-driven business models, relationships of labour and care that are highly diverse and situated. However, working within a framework of what Tsing calls 'supply-chain capitalism' (Tsing, 2009) she points out that capitalist accumulation also increasingly incorporates nonscalable relations into the global supply chain (through scalable accounting, for example). We can see something similar in the way academic publishing has been able to incorporate incredible amounts of volunteer work and underlying non-calculable relations of care into its models, so much so that these now sustain them. In this model, production need not be scalable. Think about how time spent on peer review is not standardised (and its duration varies considerably across fields, for example (Huisman and Smits, 2017)), or how reviewer experience can differ significantly. Yet these differences are eroded and standardised in peer review processes and editorial work (in a system where remuneration for these kinds of academic labour processes is either absent or based on fixed fees (often much lower than market values). 3 Or look, for example, at how the widely divergent costs of individual book production are translated into a standardised BPC. Meaningful diversity gets eroded in all kinds of different ways in these systems based on alienated labour. Here 'scalable commodities are made through the exploitation (in the natural-resource sense) of nonscalable labor and environmental relations' (Tsing, 2012: 521). The first step, then, in building nonscalability theory and identifying alternatives, Tsing argues, is to 'denaturalize scalability', which we can do by revealing the historicity of scale and economies of scale within scholarly publishing, to help us notice nonscalable phenomena, their vitality, and the links that exist between scalable projects and nonscalable worlds. As Tsing explains: 'Instead of taking scalability for granted as a necessary tool of progress, nonscalability theory attends to the work of contingency and failure. Nonscalability theory shows us scalability in action' (Tsing, 2012: 510).
Nonscalability theory, as Tsing envisions it, is about 'looking around rather than ahead', to cultivate the vulnerability to unexpected encounters (with entities, objects, disciplines). This will allow scales to arise from the relationships that inform particular projects, scenes, or events. This is in line with what we theorise here as scaling small. Here, instead of competing for the scholar or world-builder's attention by scaling up and expanding -as scalemaking projects do -the trick is to trace or make relationships between projects. This involves rethinking our knowledge or publishing practices, which have been shaped as part of an economic system that has been focused on 'remaking the world for scalability'. Nonscalability theory, in which we would include in this context the notion of scaling small, should then be perceived 'as a way to reconceptualize the world -and perhaps rebuild it' (Tsing, 2012: 524). Nonscaling is key here to the (re)making of cultural diversity and multispecies landscapes.
Following Tsing's recommendation to pay attention to how reimagining our knowledge practices requires us to pay particular attention to articulations between the scalable and the nonscalable, what is needed to enable this is first and foremost a reimagining of what academic collectivity, community, and commonality is and could be in a digital publishing environment (Adema and Moore, 2018). New forms of collaboration need to be imagined in this sense (Schneider, 2010): collaborations that won't of necessity be able to scale; that focus on nonscalable modes of valuation or gift economies; or that showcase the messiness and contradictions that arise when we do scale; and actively engages with them instead of pushing this difficulty out in the search for a more uniform and easier to scale model. It is about highlighting where scalability fails -and where nonscalable relations erupt. Reimagining the relations within the publishing system beyond a mere calculative logic focused on sustainability is essential when developing an alternative, equitable OA publishing environment, in order to enable new forms of collaboration and to redefine the future of scholarly publishing in more communal settings (Adema and Moore, 2018).

Economies of scale
Economies of scale are achieved when production goes up and costs are proportionally reduced, which in the case of the publishing industry has been most clearly achieved through the corporate growth and consolidation of publishing companies. Academic journal publishing is a highly-concentrated, heavily marketised industry with a handful of commercial organisations publishing well over half the total number of articles each year (Larivière, Haustein and Mongeon, 2015). Multinational publishers increasingly benefit from their size through economies of scale and acquisitions of smaller companies that increase the reach of their services while growing their publishing output (Posada and Chen, 2018). Although less pronounced within academic book publishing, commercialisation is still a noteworthy feature of the industry (Greco and Spendley, 2016), and there is data to suggest that concentration is an increasing trend (Guns, 2020). Yet economies of scale are a relatively recent development in scholarly publishing, as it is only after World War II that commercial publishers began to steadily increase their market share, introducing an industrial approach to the publication and dissemination process (Fyfe et al., 2017;Potts et al., 2017). More recently, digital and technological developments drastically reduced both production and distribution costs for books (e.g., see print on demand in comparison to earlier short print runs), bringing down costs per unit. It was again both the large publishers and distributors who, due to their scale, were able to make the biggest investments in digital technology or could spend most towards (outsourcing) digital printing services, and hence could make significant cost savings. Yet these developments have (in principle) also made it possible for smaller not-forprofits to make use of digital supply chains to benefit from these developments. Various OA book presses were set up recently in response to these technological opportunities. Many of them have been able to develop born-digital or digital first workflows, allowing for further cost savings, reducing costs per unit. This, it can be argued, has also reduced the importance of size to have the same (cost-saving) effects on books within this landscape (Barnes and Gatti, 2019a;Maron et al., 2016). Even though consolidation has also taken place within the HSS, 4 many university presses and independent journals and presses are still thriving and continue to play a crucial role in the publishing ecosystem (Larivière, Haustein and Mongeon, 2015;Tanner, 2016). Where large commercial publishers have been making use of economies of scale to publish across HSS fields and to make savings on production, distribution, and general overheads, while expanding their title count, small, specialised presses have relied on their strength in community building, among other things. See, for example, how scholar-led presses are often integrally connected to and embedded within a specific field or subfield, or how they represent and promote scholarship in a specific region or language group (for example, see Mattering Press, meson press, Language Science Press, and ÉSBC), thus providing these presses with an advantage when it comes to marketing and dissemination and connection to a specific scholarly community.
The rise of intermediaries or aggregators in publishing has played an important role in creating economies of scale. It is also a direct consequence of the rapid scaling of academic publishing after World War II (this scaling representing a form of outsourcing for both publishers and libraries) along with further changes to digital workflows. One consequence of this rise has been that, within book publishing, an estimated 25-50% of 'book revenues' goes to intermediaries within an increasingly complex supply chain (Barnes and Gatti, 2019a;Fisher, 2015). As the recent COPIM report emphasises, ' existing print and ebook distribution channels are difficult for new or OA publishers to engage with' (Stone et al., 2020), often acting as gatekeepers to library supply chains or global distribution channels -when distribution channels are also increasingly set up for, and controlled by, large commercial publishers. Yet the need for these expensive intermediaries within publishing is increasingly questioned.
Indeed, doing without them, as many scholar-led and not-for-profit presses do (often by necessity), can lead to significant cost savings. It is these intermediaries that partly make commercial publishing so expensive and forces it to scale, while at the same time harming diversity within the publishing ecosystem (Barnes and Gatti, 2019a;Posada and Chen, 2018). But next to relying on outsourcing and intermediaries, as Tsing has shown, economies of scale or scalability also relies on nonscalable elements (such as volunteer labour), or on public infrastructure. Within academic publishing this is clearly visible in how large commercial publishers are dependent on or scaffold upon publicly-funded institutions, from universities to libraries, and increasingly (through APCs and BPCs), funders. Within academic publishing, volunteer labour provided by scholars (when perceived as something that is part of an academic's workload 5 ), forms the publicly-funded human infrastructure the profits in academic publishing are based upon -and the nonscalable elements Tsing refers to, which are extracted for profit in supply chain capitalism.
The difference is, of course, that economies of scale are needed in an environment that is based on competition, a demand for growth, and for ever-rising profits for stakeholders. We could even argue that the development of economies of scale is integrally connected to the rise of the publishing behemoths, which has in turn led to an increasing lack of diversity in the publishing landscape and the ongoing disappearance (through mergers, acquisitions, and buy-outs) of independent publishing and scholarly communication organisations. Scale leads to consolidation. Scalability demands a market paradigm. Hence this is not a framework that we feel adequately describes the projects we are involved in. 6 For not-for-profit independent and scholar-led presses and publishing initiatives, there is no direct competition element around sales (although there is to some extent for funding), which means it has also been much easier for them to form alliances (think of Scielo, Redalyc, and AmeliCA, for example). The critique of these kinds of alliances seems to focus on how they are not large enough to compete with the large commercial presses, but why should they have to compete with them? Economies of scale are generally based on increasing output (acquisition) and increasing sales (marketing), things that often mean scaling a publishing operation. Scale is framed here in terms of competition: it is perceived as necessary in order to remain competitive. Not-for-profit alliances are formed to be more resilient, and to share services and care. We would argue that in a non-competitive environment, more important than chasing a direct profit incentive is a form of collaborative scaling that focuses on issues such as infrastructure provision, next to addressing the many hurdles small publishers encounter when it comes to interfacing with libraries. Because scale in publishing has led to these channels being made interoperable predominantly with larger corporations or intermediaries. Here the focus is more on decentralisation and retaining the independent characteristics and organisational models of individual presses, and on a flexibility that allows for a variety of business models (instead of aiming for or promoting one business model to rule them all, i.e., BPCs). We would like to put this forward as the practice and principle of scaling small.

Scaling small
The relationalities of publishing we are arguing for here focus on achieving scale across multiple organisations, with a focus on care, collectivity, and cooperation rather than competition.
In what ways will these publishing initiatives then be able to become resilient whilst, as we would call it, scaling small? Due to their size and often not-for-profit background, scholarled OA projects do face various structural constraints, from lacking skill sets and experience to insufficient market leverage. What is important to note here is that these projects tend to work according to capacity, from a few books a year to several dozens, in order to keep it manageable to the people involved. The latter is easier to achieve when there is not a profit motive (Adema and Stone, 2017). However, when taken together, in various constellations, these independent community-driven projects do have the potential to create a supportive ecosystem to sustain and build the scholarly commons. From working on individual projects to contributing to collective and collaborative ones, this will allow these projects to retain their independence and to honour their not-for-profit character, while providing a framework capable of making publishing more resilient.
As we have outlined elsewhere, two models of collaboration both characterise scholar-led presses, and, when stimulated, could help them become more resilient: firstly, a model that focuses on alliances of small independent projects within a certain sector (such as publishers) in collectives horizontally to create support structures; and, secondly, one that encompasses vertical collaboration across sectors or fields to create multi-stakeholder ecologies (Adema and Moore, 2018). Horizontal collaborations in collectives or consortia, facilitated through unions of small/independent presses, or of publishing communities taking on book series or journal projects, can provide mutual aid and logistical support, shared services, and best practices. The Open Library of Humanities (OLH), which brings together various OA journals in the humanities, is a good example of this. So too is one of the largest current collectives of presses, the Library Publishing Coalition. A US federation of research libraries involved in publishing support, it was founded in 2013 by over 60 academic and research libraries.
But beyond enabling horizontal and vertical alliances, scaling small involves the creation of infrastructures that allow many presses to thrive at multiple scales, instead of taking up a competitive model in which some presses grow stronger in expense of others, or by usurping others (Barnes and Gatti, 2019b). What is needed here is an investment in and maintenance of robust open source public infrastructure that allows this diversity to exist, instead of outsourcing the necessary digital processes to commercial entities or platforms. The similarity here between community-owned open source infrastructures and scholar-led publishing projects is that which keeps these projects alive and going: ' communities of people who care -either as developers, supporters, or as users' (Maxwell et al., 2019: 7). As the Mind the Gap report argues, the problem at the moment is that scale provides the necessary coordination and integration needed to develop publishing infrastructure in a situation where 'neither a chaotic plurality of disparate projects nor an efficiency-driven, enforced standard is itself desirable'. As they state, 'mediating between these two will require broad agreement about high-level goals, governance, and funding priorities -and perhaps some agency for integration/mediation' (Maxwell et al., 2019: 20-21). Trust remains important in scholarly communication and within community-led endeavours, and yet trust and scale do not always go hand in hand. When designing alternative open source infrastructures for the distribution of scholarly books, an assurance that these infrastructures will not be privatised or outsourced is essential. This means creating community-owned and collectively managed systems to enable shared infrastructures to scale effectively (Bilder, Lin and Neylon, 2015). It is for this reason that projects such as COPIM, which we will discuss in more depth here, aim to provide a middle ground of community coordination. Similarly, entities such as the Open and Collaborative Development Network consider how a more ethical approach to inclusive knowledge infrastructures not only considers: the tools, protocols and platforms that need to be in place in order to advance collaborative research production, but also considers socio-technical mechanisms that could deliberately allow for multiple forms of participation amongst a diverse set of actors, and actively seeks to redress power relations within a given context. (Chen et al., 2018: 11-12; see also Okune et al., 2019) Having this shared infrastructure available would also enable other presses to join more easily, enabling lower barriers for entry of new not-for-profit and scholar-led OA presses. Most scholar-led presses will not be able to scale up from their present business models to increase their volume of output, for example, but the idea of scaling small is that they should not have to. The aim is rather to keep diversity alive against the growing trend towards publishing monopolies by having several smaller publishers exist next to larger ones in non-competitive collectives. Scaling small thus involves expanding the number of presses instead of growing larger presses. As Open Book Publishers explain, scaling small is not about growing bigger themselves but instead, by developing the systems and infrastructures to support this they 'want to facilitate a more powerful expansion of OA book publishing -by facilitating the growth of more presses like ourselves, which publish OA books without charging authors' (Barnes and Gatti, 2020).
One important way in which scholar-led presses have proven to be resilient is in bringing down costs. One of the main motivations of these endeavours has been to show that it was possible to publish cheaper -and faster -than traditional publishing outlets. Many scholar-led presses, working in a non-competitive fashion, have also been very open about their finances. See the various examples of writings on the costs of publishing and of running a press that have recently been produced: to share knowledge on this front; but also to show how cost savings can be made on a small scale (e.g., Barnes and Gatti, 2020;Hall, 2015;Nordhoff, 2016Nordhoff, , 2018Oei et al., 2020). 7 Open Book Publishers provides a good example in this respect of how to bring costs down by at least a third compared to traditional commercial publishers by using alternative distribution channels. 8 Where BPCs are used, there has been an effort to reduce these, to charge them according to what authors or their institutions can afford, or to waive them completely where needed and actively to help authors find funding for their books. 9 There is similarly a strong focus on using, building, and sharing open source tools and platforms to make publishing more efficient, to reduce reliance on commercial solutions and intermediaries, and to create cost efficiencies in the system. Again there are now several efforts being developed to start bringing these together in toolkits and information platforms in order to stimulate others to set up presses. 10 And this effort towards the sharing of resources and skills characterises the larger scholar-led publishing community as a whole, where there is a focus on knowledge exchange overall, on collaboration, and mentoring of smaller and/or newer initiatives, of co-publishing, and of community and consortium forming on various levels. Scaling small is therefore also about distributing lessons learned and best practices across organisations, in relation to what Tara McPherson describes when talking about ' attempts to scale the lessons learned at Vectors over several years to broader scholarly communities', while creating 'new human and technological infrastructures for scholarly communication' (McPherson, 2010). We also see this emphasis on collaboration in experiments with publishing models: from the communal editing and publishing models favoured by Open Humanities Press and Language Science Press, to a focus on getting the community of readers more directly involved through crowd-sourcing and donations. Most importantly, perhaps, we see this in collaborations and funding arrangements with public not-for-profit institutions such as libraries and universities, who have similar motivations towards the open dissemination of scholarly content.
One other benefit of scaling small has been exactly the capability to experiment, to take risks and try out new things. This is something the larger publishing companies have been hesitant to do, being generally more risk-adverse and conservative (or cost and brand-aware). Scholar-led presses have been important trendsetters in exploring new publishing and business models for book publishing. They have been the first to fully adopt OA for books, for example. But they are also known for their cutting-edge experiments with multimodal, open, living, and processual books and publishing projects, and for exploring alternative practices and formats in HSS publishing. 11 Due to the open and reusable nature of these experiments, they tend to be taken up by other presses more easily too, where they are often developed with the aid of open source tools and platforms, instead of being bound to proprietary environments.
Scaling small also critiques the idea that publishing needs to be able to scale 'globally' instead of serving local communities, for example (such as is achieved by local or regional activist research or certain language communities (Kiesewetter, 2020;Méndez Cota, 2018)). Similarly, as Gary Hall from Open Humanities Press has argued, 'global scale', risks 'repeating and maintaining the kind of centre/periphery relationality of power we want to challenge' as part of the geopolitics of knowledge through our publishing endeavours, where a few countries in the so-called 'Global North' end up universalising what counts as valid knowledge. Here, as he states, developing in terms of collaboration and reiteration -rather than growth and expansion -can help prevent the reproduction of this state of affairs not simply by enabling us to place more emphasis on privileging non-standard contributions from others, understood geographically (i.e., in terms of the global South and East), but also in terms of BAME, LGBTQIPA, working class and other nonconforming identities. (Hall, 2019) Scaling small, or even as Hall argues, 'non-scaling', can then enable us to produce more complex, pluralistic, and antagonistically structured networks.
How then can we see scaling small work out in practice? In the next section we will explore the application of this organisational principle to three projects we have been participating in.

Building mutual reliance through the Radical Open Access Collective
The Radical Open Access Collective (ROAC) 12 was founded in 2016 as a way to build resilience between the many and various scholar-led publishers that have emerged and increased in popularity since the mid-2000s (Adema and Moore, 2017). These presses are mainly scholarled, in the sense that they are managed by working academics, often doing so as a labour of love (Pia et al., 2020) and usually with little to no remuneration or formalised, sustainable funding models (Adema and Moore, 2018;Adema and Stone, 2017). While it may not be the intention of each scholar-led press to grow -or even necessarily to publish regularly, consistently, or in the long term -there is value in building mutual reliance and sharing expertise with one another, as the collective hopes to do (Barnes, 2020).
The ROAC was launched off the back of the first Radical Open Access Conference at Coventry University in 2015. This event brought together publishers, scholars, and activists, all of them interested in exploring an ' alternative' vision for OA in the HSS; set apart from the more neoliberal version of OA focused on stimulating market competition, which has grown in dominance with the uptake of OA by commercial publishers and the growth of APC-based business models. The event led to the formation of a collective for sharing resources, expertise, and for presenting a united front in matters of advocacy. The collective currently stands at over 70 member projects representing monograph, journal, and experimental publishing projects from across the globe. The collective is designed to allow scholar-led publishers to build connections with one another, share knowledge, and to represent a distinct model of OA publishing that encourages sharing rather than competition. For example, we do this through a shared resource database, online discussions for imparting knowledge, and promoting one another's work at conferences. 13 Informal in its make-up, the ROAC seeks to stimulate community-building without being prescriptive as to how members should participate. These loose affiliations with one another help to create a space for belonging and allyship between presses that are not necessarily related by discipline, geography or situation, and thus act as a way of preventing what Tsing terms a state of 'institutionalized alienation' (Tsing, 2017: 255). From the perspective of publishing, such an institutionalised alienation reveals itself in the unsupported nature of scholar-led experiments, undervalued by universities and siloed away from their fellow practitioners. The ROAC therefore aims to counter this alienation through new proto-institutional affiliations and entanglements that arise from bringing such a disparate group together. In creating these entanglements between scholar-led presses, the ROAC works to nurture the kinds of relationships described by Tsing as being present within the 'latent commons': an often unnoticed and underdeveloped site of ally-building that can be 'mobilised in common cause' (Tsing, 2017: 253).
Commons, in general, refer to more formalised arrangements around community management of shared resources, particularly the formal rules for participation and governance (Ostrom, 2008). Yet in the latent commons, practices of mutual reliance are informal and cultivated through happenstance interactions rather than through rules or institutional structures. In fact, for Tsing, the latent commons does not 'institutionalise' well (Tsing, 2017: 255), primarily because of its ephemeral nature. So it cannot be reduced to any formal structures that presuppose a particular kind of behaviour. The ROAC is similar in this sense, due to its heterogeneity and the diversity of interactions it seeks to foster, meaning that its purpose is one of bringing disparate groups together in order to cultivate, rather than prescribe, cultures of mutual resilience. It is better understood as a 'non-institutional' institution rather than as anything particularly formalised, and this conception also has implications for our understanding of scale.
As explored above, scale erases nonscalability and works against what Tsing terms the 'transformative relationships' within a community. But it is precisely these kinds of relationships that the ROAC hopes to nurture through the simple acts of coming together, supporting one another and then sharing our knowledge with others in the collective and beyond. This is scaling small in action. Economies of scale require consistency, predictability, and modularity, whereas Tsing's argument for the latent commons is based on ephemerality and happenstance interaction. Predicated on solidarity, the ROAC's scaling small approach implies a continual negotiation between the scalable and nonscalable, the individual and collective, in the hope of supporting this transformativity without homogenisation.
Yet there are limits to the counter-hegemonic potential of the kind of large, informal, horizontally-governed space that the ROAC represents, in the same way that, for Tsing, the latent commons ' cannot redeem us' (Tsing, 2017: 255). Informal and spontaneous collectives like the ROAC are in the here and now -' amidst the trouble'. This means they are never entirely thought through, wholly democratic or formal enough to embody the kind of strategic direction or intentionality necessary to effect systemic change across scholarly communication. For example, the ROAC is limited through its overreliance on English as the main language of communication. Although it is not the ROAC's point to usher in a utopian commons for scholar-led publishing and replace the current academic publishing oligopoly, we can nevertheless look to other ways of formalised scaling that may contain more hegemonic potential in this way.

ScholarLed and the COPIM project: from latent commons to boundary commoning
The ROAC represents a predominantly informal, horizontal collective for supporting new forms of scholar-led publishing; strategically influencing the direction of scholarly communication is not its primary intention. Though valuable in itself, it is also important to explore other avenues of impacting change through strategic alliance building. Multi-stakeholder collaborations form another important strategy in making not-for-profit, independent publishing more resilient. Collaborations involving libraries, universities, funding agencies, and infrastructure providers, all with a shared interest in the public value of knowledge, are examples of this. Here, there is scope for thinking of the various not-for-profit entities within scholarly communication as potential community partners in the emerging OA commons of academic publishing, based on the kinds of vertical cross-sector collaborations described above. The aim then becomes to realign the existing resources in the system of academic publishing, and to direct them to alternative not-for-profit collaborative models.
Formed in 2018, ScholarLed 14 intends to set up and expand exactly these kinds of collaborations and relationships. A consortium of five established scholar-led academic book publishers (Mattering Press, meson press, Open Book Publishers, Open Humanities Press, and punctum books), who all are also members of the ROAC, they aim to develop new tools, workflows, infrastructures, and processes to support the consortium as well as scholar-led publishing more in general, alongside setting up new vertical alliances that will further support not-for-profit publishing more in general (Barnes and Gatti, 2019a). Unlike the ROAC, however, ScholarLed is predicated on a formal agreement between the presses in the consortium, meaning that they are bound to one another more closely as mutually reliant partners. 15 While still a nascent arrangement, the formal agreement between presses may take the form of shared legal ownership, collaborative infrastructure design, offline/online marketing and shared revenue models, allowing ScholarLed presses to exploit the scale of a larger, more visible organisation, while maintaining each press's unique identity.
This mutual reliance and close association does not make ScholarLed a homogeneous whole, although projects are more closely knit within ScholarLed than within the ROAC. ScholarLed is a collection of projects rather than an individual project, and so there is a continuous negotiation between collaboration and individuation rather than straightforward overlapping consensus. Unlike in the latent commons where interactions are based on ephemerality and spontaneity, the formality of ScholarLed necessitates a degree of structure that may foreground resilience over ephemerality, but will still seek to foster a variety of interactions, processes, and published outputs. This structure has helped ScholarLed to grow as a successful 'nonscaled' organisation. For example, it maintains, as our colleagues Lucy Barnes and Rupert Gatti show (2020), the second largest collection of published books in the Open Access in European Networks (OAPEN) library. Because of this focus on mutual reliance, ScholarLed is interested in bringing in and supporting other like-minded presses in order to strengthen the collective, and in exploring how scaling small could work with an even larger, more heterogeneous group of publishers. The latter question is being explored by ScholarLed through a focus on collaborative infrastructure design, collective governance, and shared revenue streams, most notably through the COPIM project.
Building on the work of ScholarLed and many other smaller OA publishers, the COPIM project 16 is an international partnership of researchers, universities, librarians, OA book publishers, and infrastructure providers. Funded by the Research England Development (RED) Fund and Arcadia -a charitable fund of Lisbet Rausing and Peter Baldwin -COPIM is building community-owned, open systems and infrastructures to support OA book publishing in the HSS. This includes work on revenue infrastructures, business models, dissemination systems, archiving, and experimentation, all released open source and managed through systems of governance that foreground community leadership and cooperation. COPIM intends to make it easier for small publishers to release OA monographs by building infrastructures designed not to scale existing endeavours but to promote publishing within and by constellations of local communities.
Though each is grounded in a similarly collectivising approach, we would like to suggest that the different conceptions of scale that exist between the ROAC and the COPIM project could represent an evolution from the latent commons (described above), to the more formalised practices of what Massimo De Angelis terms 'boundary commoning', or 'that type of commoning that crosses boundaries, activates and sustains relations among commons thus giving shape to commons at larger scales, pervading social spaces and intensifying the presence of commons within them' (De Angelis, 2017: 267). For De Angelis, the scale and resilience of cooperative modes of production occurs through 'structural coupling' between different projects. Here, individual projects (in our case, publishers) look for ways to formally embed themselves in one another's organisations in order to make themselves larger and mutually reliant, while still maintaining individual project identity and autonomy. It requires projects to both give and receive the fixed 'boundaries' of each other's organisations so as to create a larger whole.
From the perspective of COPIM and community-led OA publishing, boundary commoning holds great potential to scale small. It can, for example, lead to a more efficient sharing of resources, knowledge, technologies, revenue streams, and even labour arrangements, such that individual projects within COPIM see themselves as part of one another and reliant on each other's work. This could also lead to vertical alignments between other organisations of a similar outlook, such as libraries and open source technology providers. Yet boundary commoning begins from a small handful of presses working together and slowly building their work outwards by inviting other presses to participate. This process is antithetical to the commercial publishing platforms that seek economies of scale from the beginning by designing everything around the desire to expand.
The resilience of community-led publishing is therefore built through relationality rather than technology or innovative business models. Boundary commoning is a decentralised, democratic process that requires presses to think about their relationship to other projects differently as potentially transformative collaborators rather than competitors. It is important to note that, as a form of scaling small, boundary commoning does not represent a utopian vision for a harmonious world in which all presses simply collaborate without conflict. Instead, it accepts the necessity of antagonism and assumes that work is needed on governance and conflict resolution, which is where the exploratory practical work on governing the commons is so important. In this vein, COPIM is designing robust governance procedures for the project in order to set the terms for cooperation and the long-term future of infrastructural maintenance (Moore and Adema, 2020).
COPIM thus represents monograph publishing scaled 'small' through its invitation to cooperate on the development of an OA publishing ecosystem that has global reach but preserves local contexts. COPIM itself has an Anglo-US bias, being predominantly UK/US/EU based and focused in its initial application (although we are actively working on challenging this bias). This is partly due to funder requirements and partly due to the situatedness of the partners involved in the project. But COPIM's mission and aims align with similar community-led projects, cooperatives, and collectives in other regions and language-localities, most notably OPERAS, 17 AmeliCA, 18 the ROAC, CLACSO, 19 African Journals Online, 20 and Redalyc, 21 all developing not-for-profit, open, and scholar-led knowledge systems and infrastructures. Working within a higher education system so beholden to competition, COPIM thus seeks to foster collaboration with these other situated projects, aiming to slowly and carefully nurture our shared publishing futures through cooperation and the promotion of diverse cultures of knowledge. Coupled with experimentation in the latent commons of the ROAC, new forms of scaling and nonscaling begin to emerge for more resilient forms of scholar-led academic publishing.
Notes 1 See, for example, the central role care plays in Mattering Press' publishing processes: www.matteringpress.org/blog/new-forms-of-care-for-sts-books. 2 With BPCs a new set of gatekeepers, funders and institutions, will have to be navigated, and it is unclear how this dynamic will play out within the humanities especially. There are huge issues around governmentality here, and there is a risk that, similar to the subscription system, we will be creating a market for APCs (article processing charges) and BPCs based on the brand of the publisher. BPCs based on the charges that are currently being set mainly by commercial publishers, will simply not be affordable in the humanities; there is not enough money in the system (Aguado López and Becerril García, 2020;Bargheer et al., 2017;Eve et al., 2017). 3 Tsing explains in this context how standardised peer review processes within business-oriented regimes of publishing flatten the social relations around research and turn research into commodities: 'The point of academic assessment exercises, however, is to move forward the project of privatization by erasing the common space of the workshop and the social obligations in which it is entangled by turning individual papers into points. Alienation assessment privatizes and commodifies by interposing a process that is self-consciously blind to constitutive social relations' (Tsing, 2013: 21). In contrast to this, many scholar-led projects emphasise the formative aspects of peer review and perceive it as an important transformative process, providing support, care, and mentoring. See, for example: https://publicphilosophyjournal.org/about/review. 4 See, for example, Bloomsbury Academic buying Continuum, Cassell, and also T&T Clark, Berg Publishers, Methuen Drama, Arden Shakespeare, Bristol Classical Press, Fairchild Books and AVA (Eve, 2014: 36). 5 Although in practice a lot of academic service work such as this happens in overtime and is not always acknowledged formally within academic workload models. 6 In this respect Potts et al. argue that scholarly journal publishing (but we would argue a scholar-led press too) is best represented 'not as an economics of firms, markets and specialisation', where value propositions lie somewhere else, and they are arguing instead for an ' economics of team production and consumption in what [they] style "knowledge clubs"'. The focus here is on the community, on the social life of journals, and the value added by the relationalities the community weaves around a journal/press and vice versa the knowledge communities that journals or presses sustain (Potts et al., 2017). 7 For a specific discussion around overheads in this context, see: https://punctumbooks.pubpub.org/ pub/here-is-what-you-can-do-with-your-overhead-punctum-books/release/2. 8 Also see the earlier discussion around the dominance of intermediaries. 9 See, for example: www.matteringpress.org/about/funding-model-and-fees. 10 See, for example, the ROAC information platform (http://radicaloa.disruptivemedia.org. uk/resources) or the recently launched Open Access Book Network (https://hcommons. org/groups/open-access-books-network/#:~:text=The%20Open%20Access%20Books%20 Network,OPERAS%2C%20ScholarLed%20and%20SPARC%20Europe). 11 See, for example, Open Humanities Press' Living Books About Life series of editable wiki-books (www. livingbooksaboutlife.org) and Open Book Publishers Innovative Publications, incorporating ' audio/ visual material into the fabric of the text; zooming functions on selected images; using commentary functions to enable reader annotation and discussion; and creating publications from database content' (www.openbookpublishers.com/section/108/1). 12 See radicaloa.disruptivemedia.org.uk. 13 See, for example, the ScholarLed/Radical Open Access collaborative bookstand, which exists in both an offline (https://radicaloa.disruptivemedia.org.uk/collaborative-open-source-bookstandroac-flash-drives-and-postcards) and online version http://radicaloa.disruptivemedia.org.uk/latestpublications.