The annual Alliance of Digital Humanities Organizations (ADHO) conference is, for many digital humanities scholars, the one event in the academic calendar at which their research can be appreciated in all its interdisciplinary glory.[3] The slow and uneven uptake of digital research methods in many traditional humanities disciplines makes the conference a place of refuge for some; an opportunity to “nerd out” with like-minded colleagues despite otherwise evident differences in academic backgrounds. The Digital Humanities (DH) conference in this sense has adopted something of a “Big Tent” ethos (indeed, this was the official conference theme of DH2011); it is a meeting site where a wide spectrum of scholarship can be articulated and heard.

The metaphor of the “Big Tent” has been debated as a defining feature of the digital humanities itself. In one of the earliest of the “Big Tent” reflections, William Pannapacker momentarily switches metaphors to ponder, “The digital humanities may seem like a lifeboat amid the wreckage of higher education in the humanities, but it’s not large enough to hold everyone who’s still in the water”  [Pannapacker 2011b]. For Pannapaker, the question of inclusion at the DH conference is one of “scale” (however impossible that may seem to resolve). But as Ethan Watrall has noted, the issue is also one of “curation” or gatekeeping: “Whatever the perspective on the ‘Big Tent,’ the metaphor has inevitably led to debate as to who is in this ‘tent’ and who is not”  [Watrall 2016]. Watrall focuses his attention on the absence of relevant disciplines, in particular archaeology, at the annual event. Others have taken digital humanities data techniques and applied them to the conference itself to assess just how encompassing the Big Tent really is.[4]

In their article “What’s Under the Big Tent?” Scott Weingart and Nickoal Eichmann find that, although the conference has consistently grown, this expansion has not necessarily equated to a commensurate increase in the diversity of topics, disciplines, or participants [Weingart and Eichmann-Kalwara 2017]:

As Élika Ortega noted in “Zonas de Contacto: A Digital Humanities Ecology of Knowledges,” there is a “latent tension between the digital humanities that takes place in the ‘centers of gravity,’ such as the Digital Humanities annual conference…and the digital humanities that is happening outside of such centers” [Ortega 2019].

As the largest and most recognized global conference in the field of digital humanities, it is critical to reflect on what it is the conference is supposed to do, and who and what it is for? As more digital humanities conferences emerge with various national, regional, linguistic, thematic, disciplinary, methodological, and other foci, we believe that our international organization should rethink the point of the conference through the perspective of diversity, equity, inclusion and decolonization. Similarly, the conference’s stated purpose for “the benefit of its members and for the advancement of research and scholarship in the variety of disciplines and professions they represent” [ADHO n.d. “Conference”] is overdue for a reconsideration that centers justice rather than merit, equity rather than innovation, polyvocality rather than canons, differences rather than standards, and inclusion rather than gatekeeping. This article provides guidelines as a basis for future conference organizers to shape upcoming conferences.

Such a reconceptualization of the conference has the potential to strengthen the diversity of the field from its very foundation, not just as a factor of language or nationality.[5] Further, an approach to conference organizing based on these principles would also clarify what position the ADHO conference occupies in the growing (and sometimes conflicting) ecology of digital humanities conferences. To allow “Digital Cultural Empowerment” (DH2014), to encourage “Global Digital Humanities” (DH2015), to address “Digital Identities: the Past and the Future” (DH2016), to provide “Access/Accès” (DH2017), to build “Bridges/Puentes” (DH2018), to consider “Complexities” (DH2019), to meet at “Carrefours/Intersections” (DH2020): these official themes of DH conferences past all speak in support of the principles that we advocate here, but that we argue have been implemented inconsistently and with varying degrees of success. The realities of the conference processes do not always align with the themes.

The DH conference represents the good intentions of many actors (ADHO organizations, program committee, local organizers, and so forth) to welcome people to the conference and expand the digital humanities community. And yet, despite these best motivations, the conference suffers high levels of inaccessibility, language restrictions, poor distribution of knowledge and lost research. The cost of the conference is high, as is true of almost all such academic events - not just fiscally but also environmentally (all those air fuel miles...) and personally for those who give up valuable years of an academic life to deliver an event of limited or uncertain personal and community benefit. Indeed it might be argued that the real, underlying work that the conference does is to create curriculum vitae lines that recognize service rather than meet the intellectual needs of the field.

In the shadows, behind the conference tent is an opaque, hierarchically regulated, global bureaucracy which distributes decision-making in ways that make meaningful interventions for change difficult to imagine. Which bodies answer to whom? Who can assert their authority and in which contexts? When something goes wrong, who takes responsibility for answering the emails, and who addresses the issue publicly? When something goes right, who gets the credit?

We write and reflect on these questions as a group of digital humanities practitioner-scholars who have experienced the vagaries of the annual ADHO conference organization first hand. In this article, we use our front-row insights to explore the ways in which the conference is articulated, organized, and facilitated through conference protocols and processes. The conference protocols, which are bureaucratic documents established to guide the conference process, outline workflows and encode the principles of the organization. The essay generally follows the temporal flow of the conference organization process, but a number of crucial value-based themes (probably the most central themes of this essay) are woven throughout: multilingualism; equity, representation, and diversity; visibility and labor issues; and others. We start with a brief presentation of these themes, then point back to them throughout the essay as they arise in the course of conference organization.

While some aspects of the international DH conference exist because they have been thoughtfully decided by multiple committees, other elements are retained because of past practice. There is an assumption written into the conference processes that conference committee members, authors, and participants are familiar with the protocols and will be held to them equally. The reality is that few are familiar with the protocols at the level needed to have a seamless conference process. A lack of transparency in how ADHO and its conference works is also deeply felt by many digital humanities community members. Although efforts have been made to remediate a general lack of understanding of just how ADHO functions, this continues to be an issue. This essay challenges us to reflect not only on past practice but also on what should be carried forward — and what changes could be made to better reflect the values discussed here. Throughout the essay, we detail how issues such as equity around gender and representation, and multilingualism play out within the ADHO Digital Humanities conference space. We recognize that these two threads are not inclusive of all forms of equity and representation including race, ability, sexuality, ethnicity, class and all the intersectional dimensions of these that are articulated in the conference process. However, multilingualism and gender have received extensive attention from ADHO officers, conference organizers, and attendees. As such, we launch our analysis by positing that equity, particularly gender equity and multilingual representation, are not questions of proportional representation where the conference should reflect gender “balance” or “proportional” participation from the various multilingual constituencies. Rather, conference organizers should discuss what constitutes gender equity and make the results of that conversation public. Equity should, of course, be based on reparative action and restorative approaches to conference organization. These actions and approaches would require identification of all who have been harmed in previous conference work, conference attendance, and the spheres of influence created through conference processes.

ADHO, as with many other scholarly organizations, is driven by unpaid volunteers who work diligently to create a scholarly community and an international conference that reflect the current trends and advances in scholarship. This article recognizes this scholarly labor while also pointing to the structural challenges that these volunteers face. It suggests that conference organizing is not merely an act of service but is a scholarly activity that should be treated as such. From this perspective, then, we are obligated to study and critique the conference as a site of cultural production. The issues we discuss here are grounded in the digital humanities but speak to broader academic practice within academic conferences. Acknowledging these complicated challenges is the first step to rectifying them in order to create an equitable and welcoming environment for conference organizers and attendees alike.

Despite their centrality to academic communication, circulation of research, and their use as a measure of prestige within tenure and promotion dossiers, academic conferences are an understudied sociological phenomenon. Some publications about conferences are conference reports that reflect on the events and programs of a given conference; other publications consider the practical elements such as “value, management, timing, program, people, protection, scaffolds, and money” [Pedaste and Kasemets 2021, 92].[6] The little research on conferences as sites of academic knowledge production centers on four key functions: “intellectual communication, professional socialization, the reproduction of academic status hierarchies, and the legitimation of new subfields or paradigms” [Gross and Fleming 2011, 153]. The sociologists Cole and Cole pay attention to the fact that intellectual communications allocate more status and prestige to some members of the academic community than to others [Cole and Cole 1974]. Miné de Klerk points to academic conferences as sites of networking that lead to “academic career advance, to spark research collaborations, or even lead to future employment” [de Klerk 2021, 116]. Our article draws on conference organizing experience to consider how practical and organizational structures impact accessibility and diversity. Joining these two perspectives allows us to understand how knowledge production and intellectual community are rendered through conference organization.

A more recent paper by Ayesha I.T. Tulloch [Tulloch 2020] evaluating the actions and policies of conferences held by international academic societies for ecology and conservation found that while “Conferences are important for professional learning and for building academics’ reputations and networks” the quantity and quality of initiatives to promote diversity were variable, and that there are “ethical and social justice concerns when groups are excluded from participating fully.”[7] It has been suggested that monitoring performance indicators, which will allow evaluation of gender roles and inequalities in academic conferences, could tackle the underrepresentation of women [Corona-Sobrino et al 2020.]. In the wake of COVID-19, online conferences may also provide opportunities to mitigate challenges around diversity [Raby and Madden 2021]. Given that it has been shown that participation in conference organization (including session chairs and membership of the conference committee) has recognized academic prestige value [Vinkler 2017], paying attention to how conferences are organized, governed, supported, enacted, performed, promoted, and instantiated tell us much about the prestige systems that operate within particular academic domains, and the academy more generally.

The ADHO annual conference is an ideal conference to study as an example of academic conferences. It is interdisciplinary and international in nature. It is a large-scale conference that has grown rapidly throughout the past decade. It has doubled in size, now regularly attracting more than 1000 attendees. Additionally, much of the materiality of the conference including portions of its documentation from previous conferences are available online. This paper details the results of Reflection-in-Action, Reflection-on-Action, and Action Research discussions from individuals who have all held organizational roles within the Digital Humanities annual conference. Writing as program chairs, some of this article’s authors have also served in multiple roles including Local Organizers and Alliance of Digital Humanities Organizations Conference Coordinating Committee members.

Reflection-in-Action, and its counterpart, Reflection-on-Action, provides a methodological framework commonly used in Information Science and Human Computer Interaction to identify “features of the practice situation – complexity, uncertainty, instability, uniqueness and value conflict” [Schön 1983, 18]. Given the experience of the authors, a structured set of reflections upon the experience of working on the ADHO conference, and access to their contemporary notes, emails, and reports written during the conference process, were essential to the success of our task. A complementary Reflection-on-Action process, comparing and contrasting time spent volunteering when organizing past conferences, noted best (or improper) practice, and elicited the practical and theoretical knowledge gained. This helped the authors process and note both their experiences and actions, and understand and revisit underlying power structures and protocols. An Action Research recursive methodology [Stringer 2013], overlapping with collaborative self-ethnography approaches [Clapp 2017], allowed us to synthesize previous records of the decision making mechanisms for conference organization and the discussions that surround them (including decisions implemented via conference management software, emails, and minutes, many of which resided in author’s own personal archives). In practical terms, this was achieved through regular online meetings and discussions on structured topics over a six month period, with high levels of note-taking using shared online tools, interspersed with recursive co-creation, writing, and editing, using timed sprints and agreed deadlines, often setting aside periods of time to recover or locate dispersed information related to the conference that had never before been brought together, or that had not been institutionally archived. Other members of the community were contacted for records and to check details, where needed (although many chose to remain anonymous). A coordinated writing sprint brought together these various strands. Together, these collaborative, digitally mediated (and enabled) approaches elicited the history, analysis, and interpretation of the organization of the Digital Humanities annual conference, allowing full documentation, and presenting key findings which will be of use both to others in digital humanities, but also to others considering academic conference organization in other domains.

The authors of this article represent Program Committee chairs from the international ADHO conference for the most recent DH conferences. We invited all program committee chairs from DH2014 to DH2020 to join in writing the article. [8] Our experience, while not universal, can point to recent past practices, challenges, and opportunities, and, as we offer in our conclusion, possible future directions. We represent DH2014, Melissa Terras (Lausanne, Switzerland); DH2015, Deb Verhoeven (Sydney, Australia); DH2018, Élika Ortega and Glen Worthey (Mexico City, Mexico); and DH2020, Laura Estill and Jennifer Guiliano (Ottawa, Canada, canceled due to COVID-19). This article is meant to be part of an ongoing conversation that welcomes new voices and encourages reflection on our scholarly and community practices. Local Organizing committees, the Conference Coordinating Committee, and members of the digital humanities community at large are eagerly welcomed to contribute to the issues we raise here through published responses.

The international Digital Humanities conference convenes a community of digital humanities practitioners and scholars annually. The initial DH conference was held in 1989 at the University of Toronto and brought together two existing events, the 16th annual meeting of the Association for Literary and Linguistic Computing (ALLC, now the European Association for Digital Humanities, EADH) and the ninth annual meeting of International Conference on Computers and the Humanities (ICCH), which was sponsored with and by the Association for Computers and the Humanities (ACH) [ADHO n.d. “Conference”]. Subsequent conferences included participants from additional organizations that became part of the Alliance as it exists today. This international collaboration was what led, in part, to the creation of the Alliance of Digital Humanities Organizations (ADHO) in 2005 [Unsworth 2007], “an umbrella organization whose goals are to promote and support digital research and teaching across arts and humanities disciplines, drawing together humanists engaged in digital and computer-assisted research, teaching, creation, dissemination, and beyond, in all areas reflected by its diverse membership” [ADHO n.d. “About”]. In 2007, the international conference took the name Digital Humanities [year], often abbreviated to DH[year] and seen on social media as #DHyyyy.

The annual DH conference is the result of extensive labor from a changing group of people each year, who are involved in a complex committee structure that is governed by a conference organizing protocol. As Márton Demeter points out, “conference organizers and committees” are “collective agents,” “in the sense that they have an impact, or at least are capable of having a significant impact on other agents” [Demeter 2020, 43–44]. The committee structure and workflows have evolved over years, in an attempt to define — from the top down — the discipline and its labor through a “global” conference. In 2020, ADHO published a new and revised conference protocol on their site that outlines some of the workflow towards the conference [ADHO n.d. “Conference Protocol”]. The conference organization is as follows: a group of people write a proposal to host the conference; these bids are vetted by ADHO’s Conference Coordinating Committee (CCC).[9] If selected, the people who proposed to host become the Local Organizers (LO). When the location is selected, a constituent organization of ADHO is assigned as host. That organization’s leadership proposes two Program Committee (PC) chairs who have to be approved by the Conference Coordinating Committee and ADHO governing boards; until 2017 there was a single PC chair per conference. At least one of these individuals must have served previously on a Digital Humanities conference program committee in order to be confirmed as chair.

The division of labor for the conference is complicated and ADHO offers copious policy documents that attempt to describe the key requirements. The Local Organizers are in charge of the budget (including sponsorship), receptions, venues, excursions, website, and other details such as ensuring accessibility, keynote translation, etc. The Program Committee handles the creation of the call for papers, selection of keynotes (with approval by the Local Organizers),[10] peer reviewing the academic program, scheduling the academic program, and publishing the selected proceedings. While this list of tasks might seem a simple division of labor, in practice, there are many gray areas, such as coordinating the publication of the book of abstracts which is separate from a proceedings volume of full essays. Many conference elements — for instance, coordinating special interest group (SIG) workshops require the Local Organizers to work closely with the program committee chairs. Cooperation, communication, and goodwill between the Local Organizers (LO) and Program Committee (PC) chairs is essential and each conference entails subtle variations in how tasks are negotiated and allocated between them.

Further, the conference is facilitated by the “backstage” volunteer labor of hundreds of digital humanists who lend both their digital humanities expertise and their linguistic and scholarly capacities as reviewers and attendees. Little attention is given to the volunteerism that is required to organize, facilitate, and ultimately complete a full conference cycle. Committee members, chairs, reviewers, moderators, on-site staffing, etc. are each essential to realizing the conference. Yet, there is little to no attention paid to how this unpaid labor structure is enacted, inscribes inequities, alters career trajectories, and impacts personal lives.

The program committee for the annual Digital Humanities conference is composed of representatives from each of the constituent organizations, with the goal of each organization having equal representation in the development of the academic program.[11] Each association brings its own organizational interests, as does each representative. Yet, the lack of standardization of organizations within ADHO creates two intertwined problems: 1) the conflation of diverse perspectives within individual associations because of the limitations on representation and 2) the lack of perspectives beyond those held by those already reflected within a constituent organization.[12] Tension exists between individually-held perspectives and positions and those of one’s constituent organization. Is a representative's responsibility to contribute their own perspectives or is it to reflect those of one’s organization? Additionally, what exactly is one’s duty on the program committee? Is it to read and review submissions? Set policies related to the academic program? Establish the conference schedule? These and many other questions arise with the installation of the program committee and its members.

We offer a few brief examples of how these problems play out when the current PC structure mirrors ADHO’s constituent organization structure. On the one hand, one of ADHO’s constituent organizations, the European Association of Digital Humanities (EADH) spans an entire multilingual and multicultural continent yet only has one or two representatives. This flattens the differences and competing interests among the scholars inside the association that might present unique perspectives. EADH itself is comprised of associate and partner organizations, including national societies such as the Associazione Informatica Umanistica e Cultura Digita (AIUCD); one partner organization, Digital Humanities in the Nordic and Baltic Countries (DHNB), itself represents DH practitioners from multiple nations. EADH thus has many voices often represented as one. On the other hand, ADHO has multiple nationally-based societies as the Canadian Society for Digital Humanities/Société canadienne des humanités numériques (CSDH/SCHN) and the Taiwanese Association for Digital Humanities (TADH), which are smaller organizations with fewer members. While equal representation ensures a balance of power in decision-making, it does not allow for nuanced representation of the differing communities and interests that might diverge from the majority position within an organization. The lack of nuance, particularly of representation of those who are not represented by a constituent organization, undercuts the values articulated by ADHO.

Two organizations represented on the program committee are not formally based on national or continental borders: CenterNet (which represents digital humanities centers) and Humanistica (which unites francophone digital humanists from around the world). Ultimately, however, both CenterNet and Humanistica are almost always represented by well-established scholars from North America or Europe. In part, this is because the barrier to participation in CenterNet is not just having a center but it is also having the local financial resources to support CenterNet membership. For newly-established centers or those whose existence are locally precarious, the likelihood that a non-Western early-career-scholar based at a digital humanities center can both be a member of CenterNet and would be selected for this service is quite low. Humanistica offers a similar challenge in that much of the association’s membership is concentrated in North America and Europe rather than French-speaking regions outside North America and Europe. As a result, representation from non-geographically-based organizations amplifies representation of interests in those geographical regions.

We note, furthermore, that the program committee is heavily weighted towards faculty representation from established academic institutions and with already recognized professional trajectories. Part of this may lie in the issues of workload and labor where those in precarious positions or without seniority may not be able to fully participate because of the unfunded nature of the work. This is, after all, how the academy works generally. But within digital humanities, it may also be a result of being able to situate yourself within the organization previously. Those without records of participation or active long-term investment in digital humanities leadership may struggle to gather the needed support to be appointed to the position. This tendency has implications for the program committee’s ethnic representation as well, given the dire statistics on race, gender, and the academy. For example, it wasn’t until 2018 that a woman of color occupied the co-chair position in the program committee for the first time. Importantly, though, the appointment deviated from ADHO procedures. DH2018 was the “first to be organised ‘outside’ ADHO,” noted Karina van Dalem-Oskam, chair of the Steering Committee.[13] Thus Élika Ortega was appointed explicitly in a “non-voting vice chair in an advisory role,” a position created by ADHO’s Steering Committee — the first, and only such appointment. It was only later that Ortega, a junior scholar, was granted the title of “co-chair” with decision-making and voting authority equal to that of Glen Worthey, her white, male, senior colleague.[14] Additionally, the long duration of the appointments — generally 24 months minimum — makes participation by those occupying staff and non-faculty positions particularly challenging. When committee members are not appointed by the beginning of the two-year term, it impacts the functionality of the program committee. As Manfred Thaller notes in his official report, DH2016 struggled with the constituent organizations’ late-appointments of program committee members less than a year prior to the conference which meant there wasn’t sufficient time to determine collective conference priorities for the program.[15] The adoption of at-large positions is one potential way forward that might address issues of rank, diversity, and representation beyond the academy by allowing anyone within the digital humanities community the opportunity to serve. But so too would addressing the need for compensation for the work of the program committee. Solving structural imbalances in the composition of the program committee goes beyond the conference program committee and its chairs, but needs to be acknowledged in the conference’s decision making processes.

Although conference themes might not initially seem as important as logistical factors of venue, budget, and local resources that are taken into consideration as much when teams bid to host the conference, we believe them to be crucial as guiding principles in decision making processes. For example, the theme of the 2018 conference, “Bridges/Puentes,” strongly influenced the crafting of the call for papers (CfP) and the inclusion of topics such as the collaborative relationships between scholars from the Global North and the Global South; different practices and epistemologies across geopolitical realities, and others. Similarly, this theme guided the gathering of presentations over the basis of a shared topic, approach, or methodology rather than by language as it was historically done. This yielded, for example, a session in art history with presentations in English, Spanish, and Portuguese, instead of, for instance, a session in Spanish on a multitude of topics. In addition to being hosted in a large French-speaking city, DH2017’s theme “Access/Accès” also signaled the bilingualism of the conference which included not only a French language keynote, but also sessions where all the presentations were in French. Likewise, DH2020’s theme “Carrefours/Intersections” signaled the conference’s commitment to emphasizing and supporting French-language digital humanities scholarship, reflecting “the geographical and cultural heritage of Ottawa, a bilingual city in unceded Algonquin territory” [DH2020 n.d. “Conference Details”].[16] Conference themes, by virtue of being one of the first choices made about a conference, are one way that the conference signals its priorities.

While much of the content of the CfP is mandated by policy, the conference theme is handled separately. At the time of bidding for the conference, the local organizers select a theme that resonates either locally with their digital humanities community or that they feel best serves the interest of the digital humanities communities at large. That theme is shared with the executive committee of ADHO and can be subtly shifted based on their feedback. As a result, the theme is often prescribed for the Program Committee prior to its creation. This can cause tension within the Program Committee as the theme should guide the academic program; yet the representatives appointed to the Program Committee have no authority to change or revise the theme. In the best scenario, the local organizers consult with the Program Committee after their appointment on explanations and definitions of the theme that will be incorporated into the call for papers. At the end of the 2018 conference, DH2019 announced the main theme of “Complexities” which “suits the very diverse situation in Europe and DH at large, and captures the challenge to develop across regions and represent more inclusive, global perspectives on scholarship and teaching” [DH2019 n.d. “CfP English”]. Local organizers then launched a parallel theme of “Focus on Africa” [DH2019 n.d. “Focus on Africa”] that was intended to build on their existing collaborations with African scholars. While they noted that this was an effort to “help open up the conference to scholars from an area that is still very much underrepresented in the DH community,” concerns quickly arose around cost of attendance, that scholarships would only partly cover costs of travel and attendance, and that participants in the Africa workshop were somehow outside the main conference process. It did not help that the announcement was “clumsy” as program chairs noted. It also was problematic that the workshop was discussed as a “satellite” event rather than part of the conference proper and began a full week prior to the conference workshops (thereby increasing housing and meal costs for attendees who hoped to attend both). One digital humanist noted their skepticism on “how a ‘Focus on Africa’ will really empower underrepresented communities” [Kräutli 2019]. Another attendee noted via twitter that there were “few perspectives from Africa on Africa” at the conference [Kandeh 2019]. The carving-off of a portion of conference events to a separate event is not unique to 2019. Indeed, there is a track record of portions of the conference program being isolated from the main conference including being held in separate buildings by themselves (2016’s panel on inclusion and accessibility, for instance).

For DH2020, the local organizing committee and the program committee chairs co-drafted a document explaining the theme which was then discussed, modified, and endorsed by both full committees. The collaborative work on the theme helped ensure that no committee saw it as an imposition. Given the broad range of digital humanities topics and interests, the theme is a way to narrow the selection of possible keynote presenters as well as to signal to potential presenters topics that they might engage with. Most themes incorporate a set of sub-themes or topics. This can take the form of bullet-pointed lists or, as 2020 did, a set of values statements around selected sub-disciplines: of digital first nations and indigenous studies; public digital humanities; and open data. Rather than articulate all of these details in the main CfP, DH2020 elected to publish a conference themes document on the conference website that provided an explanation of each sub-discipline and its relationship to the topic of “Carrefours” [DH2020 n.d. “Conference Details”]. The program committee chose the sub-disciplinary interests and outlined capacious definitions so potential presenters could position their ongoing work within the theme. For DH2020, the program committee had planned to go forward with three themed special journal issues that would reflect each of the areas outlined, however, because of COVID-19, the Ottawa conference did not take place; as such, neither did the special issues that would have emerged from that gathering.

As the most recognized annual international conference in digital humanities, the DH conference also welcomes presentations on ongoing research regardless of its connection to the conference’s theme. This manifests a tension between submissions that adhere to the conference theme and those that do not. It is the determination of the program committee what value, if any, to assign submissions that align to the theme. For DH2020, the program committee, in consultation with the local organizers, conference coordinating committee, and executive board, chose to weight the theme as ten percent of the consideration for peer reviewing applications. The Program Committee wished to directly acknowledge the three sub-disciplinary areas as vital to the intellectual breadth and depth of the conference. Part of the rationale for this was the local appeal to scholars working on these topics; chairs also wanted the digital humanities community to consider these sub-disciplines as a coherent set of interests. The decision to weight “theme” at ten percent was ultimately frustrating for both the Program Committee and the submitters. A handful of individuals pushed back against measuring submissions in relation to theme, preferring a nominal theme that is not actually considered in relation to submissions. In some cases, the discontent stemmed from long-standing digital humanists not seeing their research interests reflected in the three sub-disciplines chosen for that year’s conference; others noted that past practice had not evaluated submissions in relation to conference themes. This concern was expressed in direct emails from participants and peer reviewers as well as in peer reviewers’ uneven deployment of the category “relation to theme” from the peer review rubric. Uneven evaluative actions were not reserved for thematic weighting alone; previous Program Committees have expressed similar sentiments about evaluation in other categories. The rubric (including the criteria on which submissions were evaluated and their weight) was circulated in advance so that everyone submitting could tailor their applications accordingly. Yet, attempts to align submission abstracts to review criteria varied widely.

In the future, the Conference Coordinating Committee, Local Organizers, and Program Committee will need to decide on the value and purpose of having a themed conference, or perhaps decide to jettison the notion of themes altogether. While the first would make it easier for submitters and the program committee to complete their work, the lack of a theme may guide the conference away from coherence. For scholars invested in emerging fields that are reflected in the sub-themes, the loss of the opportunity for the larger digital humanities community to consider their importance can be quite damaging. Ultimately, conference themes such as DH2011’s “Big Tent Digital Humanities,” can spawn debate, affect public discourse beyond the conference itself, and may have long lasting reverberations.

We recommend that program committee chairs (perhaps in concert with the conference coordinating committee and the local organizing committee) establish and publish their conference’s guiding principles, which can then be incorporated into the theme and CfP. We further recommend that program committee chairs publicly share the conference’s anticipated and actual outcomes based on the established theme. Ultimately, we view conference themes as an opening through which ADHO and annual organizers can reach out to scholars and scholarship that may have been overlooked or outright ignored.

A vital part of conference planning is the solicitation of participation to potential attendees, both those who are existing members of the association and those who might join in order to participate in the conference. Generally, membership enrollment spikes for organizations who are supporting the conference occuring in their geographical locale. The CfP serves as the first element of outreach to participants. It guides potential attendees in understanding both the mechanics of the conference (dates, submission formats, etc.) as well as the placement of the conference within the ecosystem of disciplinary conventions. These practices are not just ethical questions related to submitting one’s own original work and participating in single-anonymous, double-anonymous, or open peer review;[17] they are also statements that communicate values around diversity, inclusion, and accessibility. Does the conference welcome submissions in multiple languages, or formats? Does the call include statements related to codes of conduct and guidance on the accessibility of presentation materials? And, as importantly, does the call for papers clearly articulate how submission decisions will be reached? Do submissions need to comply with criteria such as racial, gender, representation, or status balance on panels or forums? Does it welcome submissions by those outside academic ranks? Each of these questions serves to broaden or limit the potential pool of participants.

At the annual international digital humanities conference, the call for papers process is prescribed and supervised by ADHO’s Conference Coordinating Committee (CCC). The CCC provides to every program committee a template that outlines not just what content should be included in the CfP but also the placement of elements within the CfP in relation to one another. It is guided by the “Conference Protocols” which were created in order to standardize conference procedures from year to year and also to help organizers, who come to the position new each year. The conference is organized on a rolling basis. As such, at any given time, there are three conference planning groups running: the current conference process, the past conference teams as they close out their activities such as publishing proceedings, and future conference groups who are learning about the activities they will need to participate in. While there is overlap in representation from the same organizations, very rarely is there overlap with the individual members assigned. As such, conference protocols are designed to reduce potential errors or issues, both known and unknown, for organizers, committee members, and participants as they move through one of the three stages. Manfred Thaller noted that “the relationship between [the program committee and the conference coordinating committee] has been rather vaguely defined” [Thaller 2016] , particularly in relation to when the PC needs to seek approval from the CCC. He further highlights the inconsistency of involvement of the chair of the conference coordinating committee in supervising program committee work. In 2019, Diane Jakacki and Brian Croxall, chairs of the Conference Coordinating Committee, undertook the task of rewriting the conference protocols: the new protocols were to be applied to the 2021 conference (which, due to COVID-19, has been pushed back a year); yet, even with ongoing revisions to the protocols, the ongoing debate over the relationship between the conference coordinating committee, the local organizing committee, and the program committee continues. The kinds of questions covered by the conference protocols include: What types of presentations will the conference have? Is there a conference theme or set of topics that will guide conference interests? How will submissions be reviewed? By whom will the program ultimately be decided? While these may seem simple mechanistic questions that are formulaic, the reality is that each of these questions is situated within a set of conventions that implicitly and explicitly communicate disciplinary standards and value-systems. Every member who participates in the process brings with them their own disciplinary and personal predispositions to organizing the conference.

Any changes to the template, be it language of publication, variation to submission deadline dates, the introduction of new formats or criteria for submission, or the excision of elements from the template, have to be reviewed and approved by the CCC who generally consult with the Executive officers. The ADHO CfP is created each year by the program committee, whose chairs and members change for every conference. Each program committee draws on previous years’ CfPs: as you can see from the appended calls from 2015-2020, some sections were retained over the years. This was a purposeful policy on ADHO’s part in order to minimize the labor of translating the call into multiple languages; however, it also constrained program committees from innovation, acknowledging the rapid changes in the field, or even addressing existing challenges. So, for example in 2015, the program committee was expressly forbidden from varying the CfP in their content or structure. Translation of new material would be too costly and time-consuming; this despite the fact that ADHO does not fund translation activities and that program chairs often must use their own personal resources to fund translation. Coupled with this, in 2015, even without variation to the CfP, there were no translators available for some languages. Translation may also be limited at the request of local organizers. In 2019, local organizers declined to have the call for papers translated into Dutch claiming that “most of the academic communication in the Netherlands happens in English” [Ciotti and Pierazzo 2019]. While this was likely correct, the question of whether the lack of translation to Dutch limited potential contributions from those outside the English-speaking academic community remained unanswered.

It may not be immediately obvious, but important opportunities to enact any kind of changes to the conference program, including adherence to the theme, linguistic, regional, and disciplinary diversity, rely on the peer review process.[18] Peer review offers potential that is matched with unexpected risks for failure. Conference protocols, developed by ADHO’s Conference Coordinating Committee and used for planning the event, dictates that program committees are unable to implement changes in the peer review process without approval. This has included in previous iterations of the protocols limiting the expansion of the reviewer pool to a biannual process despite need for new reviewers, as well as mandating a process of review bidding that was implemented to encourage participation by reviewers, but may have had unintended negative consequences, such as favoritism. For example, most of the procedural changes implemented for DH2018 in Mexico City (e.g., expanding the reviewer pool during an “off-cycle” year; abandoning the review bidding phase; and moving to double-anonymous peer review) were motivated specifically by the desire to make space for scholars from Latin America in the program. The existing approved review pool had limited Spanish-speaking reviewers so an expansion of the pool was clearly in order; Latin American scholars who had previously submitted to the conference had received hostile or uncollegial reviews. Moving to an expanded pool of reviewers and the double-anonymous review process was intended to alleviate those issues.

Complicating all peer review processes for the DH conferences is the use of the ConfTool management system, which utilizes both a manual and automatic review assignment process [ConfTool n.d. “Automatic Review”]. ConfTool is a commercial service that ADHO licenses each year. Every year, the previous year’s ConfTool database is imported into a new ConfTool instance, a functionality that has advantages for organizers and participants. Past participants don’t need to create a new login account or state their review expertise topics; organizers have a ready-made mailing list and review information already available to the program committee. Carrying over this database, however, also reiterates outdated user information (such as the use of deadnames, that is, original birth names after an individual has altered their name) as well as biases (listing some subfields and not others) that may have been incorporated in previous conference processes. For example, there may be missing expertise fields, scholars may not update their expertise to align to new areas of work, or there may be reviewers who remain within the reviewer pool despite issues around code of conduct. Noting below how ConfTool shapes the organization and the DH conference, we are also aware that every conference management tool will have its own issues. However, there has been no systematic evaluation of how ConfTool structures affect decision-making, organization, labor issues, and other crucial conference functions.

In ConfTool, once all submissions have been received with keywords selected and all reviewers have been marked with reviewer status, the automatic assignment system identifies conflicts of interests between submissions and potential reviewers. When conflicts have been identified, the algorithm tries to find “the best reviewers for each paper (so not the best papers for each reviewer)” [ConfTool n.d. “Automatic Review”]. In some years, this includes considering reviewer preferences (known as the “bidding process”) for particular review assignments. Most years, though, the algorithm moves directly to “matching topics according to the areas of expertise” (counting down from the highest number of matches to one, e.g., five, four, three, two, one matches between areas of expertise and topics selected by the authors). Finally, the algorithm attempts to assign to those with the fewest number of reviews assigned to a given reviewer. While this sounds logical, the algorithm is limited in three key ways. First, the algorithm does not accommodate submissions and reviewer expertise outside of the English language. Thus, you cannot assign Spanish-language submissions, for example, only to reviewers who are fluent in Spanish. Instead, you must manually match those with multilingual expertise to those submissions that are in that language adding significant labor to the work of program chairs. Second, complicating manual assignment is that ConfTool has no specific mechanism for differentiating the language of presentation from the content of the submission. As a result, program chairs have to incorporate language proficiency in the default user profiles and then manually match that information to the topical fields related to language. Because many authors either fail to mark the language of presentation, assume that the system will automatically identify multilingual submissions, or confuse the topic of research (on Mexico, for example) with the desire to submit and present in Spanish, program chairs either must review all submissions to ensure the correct languages are selected or, as most program chairs have done, manually assign all non-English submissions first outside the algorithmic assignment process. While in some languages there may only be a handful of submissions, depending on the year and location the number of non-English submissions can be quite large. This leads to the third problem which is that there is no way to weight or sequence the algorithm processing of topical keywords. Program chairs cannot prioritize matching linguistic proficiencies first before topics such as discipline or geography. As a result, the algorithm may identify a high match of four or five topics and none of them relate to linguistic ability resulting in a reviewer being unable to review a submission. While individually these issues can be resolved through creative solutions and human engineering, the underlying issue of the algorithm illustrates an English bias that undermines diversity efforts. Given the international nature of the conference, the inability of the ConfTool system to accommodate multilingualism and complexity in reviewer assignment and submission information is quite problematic. As Elena Pierazzo and Fabio Ciotti noted in their 2019 chairs report, “no automatic system will solve all the problems, nor an absolute and mechanic reliance on external peer review is acceptable in all cases” [Ciotti and Pierazzo 2019].

The manual assignment of reviewers and the expansion of the multilingual reviewer pool are invisible to most but are key for the success of the review process. These two efforts can have positive repercussions in the acceptance of proposals that under different review bodies with particular merit standards might not be ranked positively, in the overall composition of the program, and ultimately in the overall proceedings of the conference. Difficulties to enact changes such as these include making proposals to the CCC that justify a deviation from the protocols and usages at the time. That is in addition to the work of reaching out to reviewers, vetting them, and maintaining good communication. These are challenges for every conference but are especially pressing for conferences being hosted for the first time by a member association. DH2018, the first to take place in the Global South, was especially burdened as missteps in the conference process would impact both future hosting abilities as well as the potential growth of local membership.

For many years, the standard modality of peer review for the international DH conference was single-anonymous: anonymous reviewers were able to see the names of proposal authors. This practice not only was fairly non-standard in academia, but also entailed a lack of balance and equity between reviewers and submission authors [Okike et al 2016]; [Tompkins, Zhang, and Heavlin 2017]. Although at various times, single-anonymous reviewing was questioned in the ADHO community, there was considerable resistance (and perhaps marginal or marginalized motivation) to change it. Because participants in the conference come from varied institutional, geographic, and disciplinary backgrounds, however, many people bring different expectations for peer review to the conference itself.

Changes to the single-anonymous model have been used to try to mitigate these concerns, which were widely, if too often quietly, acknowledged by conference organizers, delegates, and other practitioners. Uncollegial and outright rude review tone, implicit bias, and inappropriate deference to “stardom” were known problems for years. In DH2018, for example, the shift to double-anonymous peer review helped prevent both the implicit negative biases that might arise from a reviewer reacting to a linguistically or culturally coded author’s name or from an institutional affiliation. It also addressed an unconscious (or perhaps conscious) positive bias in response to seeing a very well-known author. Some of these biases were further exacerbated by the review bidding phase where potential reviewers had the chance to choose what and whose work to evaluate (this practice between 2013 and 2017 has been abandoned). The problem of uncollegial tone in reviews was sadly not resolved by the double-anonymous approach for 2018. However, double-anonymous peer review did have an impact in closing the scoring gaps between well-known scholars and newcomers to the conference for both DH2018 and DH2019. Fabio Ciotti and Elena Pierazzo noted that double anonymous “seems to lower the authority bias and help younger and less known scholars to get good scores.” [Ciotti and Pierazzo 2019] DH2019 did, however, experience issues around the anonymity of submissions. Some authors failed to properly anonymize their contributions while others were overly-anonymized by limiting access to websites and other supporting citations that might strengthen their contributions [Ciotti and Pierazzo 2019]. DH2020 enacted open peer review where both the reviewer and submitter names were known to each other and to the program committee (but reviews were not publicly posted). DH2020’s open peer review was met with both criticism and praise. Some participants in the digital humanities conference are committed to only undertaking open peer review, whereas others can be skeptical of its scholarly rigor. Future program committees will need to judiciously choose their peer review modality (open, single-anonymous, or double-anonymous); and, regardless of the style of peer reviewing selected, will need to work to ensure that the review process does not simply reinforce existing biases and axes of privilege that cut across the conference and community.

The selection process and the making of the program is the stage of the conference where the Program Committee exerts the most power and situates a year’s conference in regards to the field. Although the selection process might seem to be a mechanical one where a “cutoff point” in scoring separates accepted proposals from those not accepted, in reality, this exercise involves a large degree of give and take. There are a variety of factors that must necessarily guide the process including the number of submissions in various presentation categories (poster, panel, long or short paper, etc.); physical limitations of the venue; temporal limitations of the conference calendar; diversity of participants (discipline and approach; career stage; geographical, linguistic, and cultural background). The merit and quality of a proposal may be perceived to be the deciding factors, as though the selection process was a competition and acceptance to the conference was the prize, one that could signal membership to the digital humanities community and room in the “Big Tent.” Complicating the question of acceptance is the rapidly increasing scale of the conference which has seen exponential growth in both the number of submissions and the number of individuals involved in the conference process through reviewing. DH2015, for example, as the table below outlines, saw 426 submissions with 371 approved reviewers. DH2019, just four years later, saw 893 submissions with an active reviewer pool of 661 individuals.[19]

A rejection notification is sometimes met with push-back, privately and/or publicly, arguing for the quality of a proposal, the precedent of having been accepted in previous years, or, from another stance, the gatekeeping of the field and the lack of diversity. As has been proven, merit in the academy is not an equitable measurement of success or rewards, but rather a skewed one where resources are allocated to those who already have them. Thus, an assumption of the fairness of rewarding merit can only lead to a feedback loop where success is granted only (or mostly) to those who are already successful [Weingart 2016]. Should merit then be centered as the deciding factor of acceptance? We contend that it shouldn’t. Instead, program committees should prioritize a deliberate curatorial approach: a local language, under-recognized academic approaches, justice and equity, etc. Conference themes (discussed above) might assist in that effort; however, without thematic alignment being represented in the review criteria, thematic efforts might be ignored. A look at the history of the conference offers examples of both.

In 2018, for example, the program co-chairs centered the location of the conference, which afforded many digital humanists in Latin America the possibility of attending due to proximity, fewer (or no) visa requirements, and lower costs, in their curation of the program. Rather than relying on the metrics of merit (scoring) alone, the program chairs focused on ensuring that the percentage of presentations accepted from practitioners based in Latin America roughly matched the percentage of submissions initially received from the region. This labor was strongly facilitated by the work already done by the reviewers who not only had various linguistic and disciplinary backgrounds, but also, crucially, came from varied academic cultures so that a single notion of merit, quality, or value did not dominate. Perhaps unsurprisingly, the single notion that dominated prior efforts in addressing inequity in review was the centering of tenure-track and tenured faculty who sought to uphold the privilege of degree, access to publication venues, and the reification of citational privilege. As the 2018 chairs learned from this example, evaluation must not only deal with linguistic competency, but must also consider cultural, contextual, and citational practices fairly. We suggest that there is no innate, universal knowledge that all scholars should have; rather, we are suggesting that the notion of what is appropriate academic practice varies greatly and can be weaponized against those who are not engaged with North American and European digital humanities as led by the tenured and those on the tenure track.

Although protocols outline that the program committee determines conference acceptance, in practicality members of the ADHO Executive have had significant interest in acceptance rates. For instance, in 2020, ADHO provided suggested baseline thresholds for acceptance and rejection without examining the submissions or their reviews; in 2015, there was significant tension around acceptance rates given the venue location (Sydney, Australia). Further, there is a disconnect between the desire by ADHO to maintain a stance of exclusivity in relationship to acceptance and rejection rates and the inverse desire for the conference to serve as a gathering opportunity for the entirety of the field that generates revenue. There are significant financial implications to exclusivity as it limits the potential income of the conference. As conference income is dispersed among the member organizations, choices to cap the number of presenters also limit the amount of disbursements to organizations in the year following. This can then directly impact the constituent organizations and their activities.

A centering of what might seem priorities alternative to merit for the selection and crafting of the program has important ramifications for the scholarly dialogue not just across disciplines, methodologies, but also academic cultures. All conferences are exclusionary in some way. But awareness of that exclusion needs to be not only attended to but also openly addressed where possible. Successive Digital Humanities conferences have consistently attempted to just add more concurrent sessions as if that would address issues of exclusion and merit; yet, in our experience and analysis, additional sessions do not necessarily translate into a more diverse conference program, rather it provides more opportunities to those already centered within the field. Ultimately, centering specific disciplines, themes, or locales can offer new points of departure for our initial questions, what is the conference supposed to do, and what is it for?

The word keynote is derived from the practice of playing a note at the outset of a musical performance in order to establish the key [OED 2021a]. Typically, this would be the lowest note (the tonic) of the musical scale. From the early 1900s, the noun has been used as a verb to describe presentations which set the prevailing tone or essential ideas of a meeting [OED 2021b]; [Kelly 2016]. Built into the very idea of the keyote then, is a sense of establishing or leading the development of harmonious relationships, of providing the conditions for collaborative expression at scale.

Academic conferences carry this meaning into the organization of keynotes or plenary speakers. A keynote speaker is expected to represent leadership in a given field, signaling both evident accomplishment and simultaneously pointing to the directions that the discipline (and more specifically the conference) might follow. Keynotes are a matter-of-course event within the format of many intellectual meetings. Yet there have been few analyses undertaken of either the selection or content of keynote speeches, and how they both signify and enact prestige, power, hierarchy, and direction of travel for the discipline.[20] In complex, manifold fields such as the Digital Humanities there is a tradition of inviting two to three (and in some years up to five) keynote speakers to cover developments in the many distinct strands of research undertaken by the scholarly community. Typically, a keynote address is delivered by a soloist and there is no competing program event to distract conference attendees. The presenter is expected to speak authoritatively on their area of expertise for forty to forty-five minutes with a short opportunity for questions from the floor.

The selection of ADHO keynote speakers is one of the first and one of the trickiest tasks of the conference programming committee after CfP creation for a number of reasons. In order to attract keynote speakers the conference covers travel, accommodation and registration fees. As there is no guarantee that the speaker(s) desired most by the Program Committee will agree, the complex logistics of these arrangements require invitations to be sent early. The announcement of keynote speakers is crucial in promoting the conference and attracting attendees — the earlier they are confirmed the higher the potential impact on the budget that pays their way by encouraging registration of potential attendees. The proposed keynotes are then required to satisfy both the expectations of the programming committee (to be of sufficient academic standing in relation to the conference themes and lines of enquiry) and the local organizers (to be sufficiently recognizable to market the conference and generate registrations). This can give rise to delicate, and not always harmonious, negotiations between the Chair(s) of the Programming Committee and the Local Organizer(s). The intensity of these negotiations are directly proportional to the stakes for the organizers which extend beyond the terms of the conference itself. For example, there are implicit career and professional network benefits to conference organizers in a position to offer keynotes to more senior and influential figures in the field. Often organizers garner individual or small group type activities with speakers that can foster collaboration or reciprocal invitations. There can also be a sense of obligation to invite specific individuals based at the university hosting the conference or located in the country in which it takes place. Not inviting a leading figure at one’s university or national network can have professional consequences. There are also professional opportunities for invited keynotes that they might otherwise not have. These types of presentations for tenure, or promotion, are taken as an indicator of value and prestige within the academic system. Often, speakers will then be invited to give additional talks at other conferences and institutions by those in attendance as a result of having seen them speak.[21] There is, then, a significant amount of power in both the choosing of academic speakers and also for those who accept to speak: the allocation of this power dynamic has hitherto been underexplored in digital humanities.

The strenuous contestation over keynote selection is not surprising given the symbolic weight of the keynote as an exemplar of how the digital humanities, and academia more broadly framed, is seen to define “merit” in a value judgment that has real-life implications for the CVs and career progressions of those offered keynote slots and the networks and reputational affordances for those that select them. This extends to the awarding of annual prizes at the Digital Humanities conference as well. These power issues often have a gendered aspect.[22] For example, the 2013 conference in Nebraska was sharply rebuked by attendees when almost all the awards and keynote addresses were offered to those identifying publicly as men. It has been noted that although, in binary terms, there is near gender parity across attendees of Digital Humanities conferences, there is a far higher concentration of men presenting standard papers [Eichmann-Kalwara, Jorgensen, and Weingart 2018]. But, given that the conference is made up of parallel sessions, each conference participant can only attend a select number of these and it is impossible to personally experience the power disparity at full scale. Keynote and plenaries are a different matter since there are no competing events and all conference attendees can be in the room for the one session. On the first day of proceedings at the 2015 conference, in which only men were seen on stage (nine of them, one after another), a vigorous backchannel twitter discussion arose in response to what was described as the “parade of the patriarchs” (See Figure 1).

We used a Content Analysis recursive methodology [Krippendorff 2018] to look at gender distribution in the keynotes of Digital Humanities conferences. For 31 years of ADHO conferences (1990-2020 inclusive) we analyzed the gender of invited keynote speakers (assessed by pronouns used by the speakers themselves in the published record). To the best of our knowledge there were no explicit non-binary keynote speakers, an omission which itself demands reflection and redress. There were 73 keynote speaker slots (counting repeating individuals and “duets” in which two speakers filled a slot): these went to men 51 times, and women 22 times. There were 67 unique speakers (47 men, and 20 women). 5 individuals had been asked to deliver a keynote multiple times: Burrows (3), Drucker (2), Queman (2), Aguera (2) and McCarty (2). There were 12 conferences with men-only keynotes: 3 conferences with women-only keynotes, and 7 conferences where there were equal numbers of keynotes. If we were to aim to achieve parity, assuming that there will be only two unique keynote speaker slots at future conferences, we would need to program women-only keynotes for the next 15 years of conferences, until 2035. This deficit persists despite the fact that gender disparities in keynote addresses have improved in more recent times: in the first 8 years of conferences in our data only men gave keynotes. Since 2012 there has been at least one woman keynote speaker at each of the conferences. In some instances, this opportunity has been created by expanding the number of keynotes presented rather than not including men speakers. In other words, there haven't been fewer men. The last all-men keynote line-up was 2011 which had four men keynote speakers (including a closing keynote presented jointly by two men). But, as noted, the pace of change is not enough to deliver on the promise of parity for some time to come.

The homophily of keynote invitations is further underscored in an analysis of the institutional affiliations of speakers. Of the 73 keynote speaking roles — with some keynote sessions involving two speakers — 72 reported an institutional (almost always academic) affiliation. Over a third (25) worked for European institutions and it should be noted that 50% (36) were located at universities and research institutes in the USA and Canada. The entire, vast and various, rest of the world (ROW) contributed only 11 speakers. Many of these 11 were programmed in the most recent three conferences. To put that a different way: in the first 22 years of the conference (1990-2011) there were only two keynote presentations in which the speaker was affiliated with an institution outside Europe and North America — and both those presentations were given by the same man (John Burrows from the University of Newcastle in Australia). Four years later, in 2015, the conference was finally located, for the first time, outside either Europe or USA/Canada in Sydney (Australia). By this time (including the 2015 Sydney conference) the accumulated ROW keynotes numbered seven, three of which had been delivered by the one man, again — John Burrows. Men are uniformly overrepresented as a proportion of speakers in this geo-locational analysis — the same percentage (72%) of the speaking roles in the Europe/UK and USA/Canada cohorts respectively were filled by men. As a proportion of the ROW women make up 36% of the total speaker slots which expands to 44% if we adjust for “Burrowing,” i.e. the effect of repeated keynote appearances by John Burrows. Based on an analysis of the aggregate numbers alone, if you want to secure selection as a keynote presenter at an ADHO conference, it would help to be a white man, ideally based at a North American institution and preferably named John (or a variation thereof).

The practice of equity and diversity is something to be publicly enacted, modeled, and celebrated: to be put “center stage” (both literally and figuratively). Some conference organizers but to our mind, not nearly enough have worked consciously to ensure that people from historically under-represented groups occupied prominent positions at the conference, including the delivery of keynotes. Most often, and despite the best efforts of these individual conference chairs, the program has not delivered on rudimentary equity and diversity baselines, notably in terms of keynote speakers. We must ask the community: why?

In part, the selection of keynotes often relies on issues of academic prestige and rank. Senior scholars who are overwhelmingly male and white within digital humanities contexts have, as we’ve demonstrated, been rewarded with keynote opportunities. There is a general reluctance for keynotes to be assigned to those who aren’t well-known within national or global digital humanities. This then reproduces their notoriety. Problematizing the selection of keynotes is the requirement that local organizers have input on keynote selection. This can manifest in conflicts over whether a potential speaker will be appealing enough to encourage people to register for the conference. It can also manifest in local organizers strongly desiring a given speaker to recognize someone from their own national contexts. Neither program committees nor local organizers are without blame in this dynamic; the conflict between them is baked in through the conference protocols where the local organizers budget for the keynote which leads them to further want input into the selection. While the keynote is technically part of the academic program, the “sale-ability” of the person is of utmost concern as there is a perception that the keynote might attract participants who otherwise might not attend.

Equity is both an outcome and a process. Without greater transparency and accountability, justice for underrepresented communities will not be delivered by the ADHO conference. It is not without some small measure of irony that a Digital Humanities conference system is unable to apply even a small number of data-driven approaches to implementing open decision making and transparent answerability to the digital humanities community. In addition to the regional and linguistic profile of speakers, we support an intentional effort to bring to the forefront practices that are not necessarily in the current core of the field. To our minds, some of the most memorable keynote speeches have shared an interest in presenting other ways of seeing, understanding, and computing the world. They have also had in common an emphasis on how digital technologies can be used for ethnic, cultural, epistemological, and sometimes physical survival. We therefore suggest that it should be a prerequisite in conference organization protocols to pledge to have a multiplicity of genders across keynote invitations, with weighting towards non-male members of academic staff, to give them access to the prestige mechanisms which exist within our scholarly societies, that will go on to impact their careers.

Historically, English has been the default working language of The Alliance of Digital Humanities Organizations and, therefore, of the conference and of the program committee. As Quinn Dombrowski and Patrick Burns note, the field of digital humanities itself has an “English problem” [Dombrowski and Burns 2022].[23] Acknowledging the hegemony of English and the academic cultures it represents, ADHO has for a long time had a stated purpose of fostering and increasing the representation of ethnic, racial, linguistic, gender, and national digital humanists. The creation of the Committee on Multi-Lingualism & Multi-Culturalism committee (MLMC) in 2005 is evidence of this aspiration [ADHO n.d. “Multi-Lingualism & Multi-Culturalism Committee”]. For ADHO’s annual DH conference, in particular, the Conference Code of Conduct directs us to “work actively toward the creation of a more diverse, welcoming, and inclusive global community of digital humanities scholars and practitioners” [ADHO n.d. “Conference Code of Conduct”]. The creation of MLMC and a lack of other bodies tasked with overseeing equity in terms of gender, ability, career stage, etc. signals how linguistic diversity and multiculturalism have been used as shorthand for “general” diversity. At its best, if actually coupled with multiculturalism, linguistic diversity could serve as a proxy for regional and epistemological diversity, but there has already been a struggle to enact that alignment.

For one, the centrality of these principles and the concrete ways they are applied has rested with the conference organizing and program committees, which change personnel and priorities on a yearly basis. Moreover, MLMC has ex officio representation on these committees but has no budget and no direct responsibility or authority for the actions of the committee — or of ADHO in general. Indeed, the responsibilities of the MLMC for the conference have been narrowly focused on coordinating the translation of the CfP from English to ADHO’s other four official languages: French, Spanish, German, and Italian. Translations into other languages – Polish in Krakow 2016 [DH2016 n.d.], Portuguese in 2018 [DH2018 n.d.], and Anishinàbemiwin/Algonquin in Ottawa 2020 [DH2020 n.d. “CfP: Anishinabeg”] – have often fallen onto program committee chairs or local organizers to do themselves or to obtain funding to pay for translation services. The contradictory aspiration of diversity through multilingualism, the lack of influence from MLMC, and the annually diverging approaches of conference organizers have resulted not just on radically uneven results from one year to another but, worse, on disappointing measures.

Despite procedures that could be expected to make a difference, as happened in DH2014 when the Call for Papers was published in twenty-three languages, multilingual participation was extremely low. As Martin Grandjean wrote in his report on the result of that appeal, submissions to the conference were in only six languages other than English, resulting in just 4.1% of all presentations being delivered in French, Spanish, Italian, or German [Grandjean 2014]. In this scenario, the aspirations for a multilingual conference failed to materialize, and also added additional overhead to the peer review process, including identifying and gathering enough reviews in minority languages. While those less knowledgeable might assume that the lack of participation was a result of an absence of digital humanities activities, or worse a deficiency in quality, in areas working in those languages, the reality is that the lack of submissions has likely been a result of systemic obstacles to multi-lingual participation. This includes the patent inadequacy of multilingual affordances in the conference management tool, the lack of resources allocated to translation efforts, the scarcity of identified peer reviewers willing or able to evaluate non-English abstracts, and a lack of outreach to multilingual communities who engage in digital humanities practices but do not strictly define themselves as digital humanists.

More unfortunate, however, has been a lack of tolerance for presentations not in English from conference delegates. Accounts of conference rooms being partially vacated when presentations were offered in Spanish, French, German, or Italian have not been rare. This has occurred despite the origination and support for the DH Whispers campaign, implemented in 2014 by Alex Gil and Élika Ortega. Attendees who were willing to help out translating to or from any language being used at the conference were able to wear a pin that demarcated their willingness to translate in support of presenters and audience members [Ortega 2014]. That work was incorporated into the Global Outlook: Digital Humanities’ Translation Toolkit, a set of recommendations for translation at conferences [GO:DH 2016]. The limitations of these efforts, though, reside in the unfunded mandate of this work as well as the linguistic limitations associated with the community itself [Spence and Brandao 2021].

Despite having to pay for translations costs, working on translations themselves, and having to present and answer questions in their second or third language, non-native English speakers have felt coerced into submitting proposals and delivering papers in English in order to receive the attention of their colleagues. This complex set of organizational and cultural attitudes towards multilingual conferencing poses challenges that despite concerted efforts have not always panned out. A review of conference acceptance data across years shows that while in 2018 (Mexico City), about 18% of the accepted papers and posters were in a language other than English, this was nearly double the next-highest levels of conference multilinguality: there were about 9% non-English proposal accepted to the two preceding conferences (Krákow in 2016 and Montréal in 2017), but prior to 2016, the percentage of non-English papers was most often 0% or 1%, with a local high of just 4% in Lausanne in 2014. We have to look all the way back to the 2006 conference in Paris (which predates both ADHO and its official “DH” conference designation) to find similar linguistic diversity: it included about 10% non-English papers. And although the 2016-2018 percentages on multilingualism may appear to represent a strongly positive current trend, in fact these statistics have since returned to just above their previous extreme low: the 2019 (Utrecht) conference included about 2% non-English presentations, and 2020 conference (to have been in Ottawa, but canceled due to COVID-19) included about 3% non-English acceptances despite a strong local French-speaking digital humanities community.

The most recent four years of the DH conference (2017–2020) offer case studies to examine the shifts in the implementation of the stated importance of linguistic diversity. DH2017 and DH2018, celebrated in Montreal and Mexico City respectively, were both officially bilingual. This included bilingual keynote presentations as well as panels that shared a topic or methodology but included papers in both languages. This was likely facilitated by having members in the program committees and/or the local organizers who were fluent in both languages and belonged to scholarly communities working across English-French and English-Spanish. We believe this may have encouraged submissions by reaching out to immediate networks. Conversely, the CfP for DH2019 in Utrecht explicitly stated that “The primary language of the conference will be English, but we warmly invite proposals written in other languages for which we have a sufficient pool of peer reviewers (German, Italian, French and Spanish)” [DH2019 n.d. “CfP English”]. With a less active approach to multilingualism, expectedly, DH2019 saw very low submissions in languages other than English. This was particularly pronounced as the accepted program was the largest to date in the history of the Digital Humanities conference. In contrast, DH2020 in Ottawa explicitly solicited submissions in all five required ADHO languages and centered multiculturalism rather than linguistic diversity, bringing to the forefront native, indigenous and decolonial studies, while also acknowledging Ottawa’s official bilingualism. Yet, there were no submissions in Anishinàbemiwin/Algonquin and few in French. As a result, “the primary language” defaulted to English despite efforts to welcome non-English submissions.

As these examples suggest, despite the establishment of the MLMC and being a central concern within ADHO, the actual practice of multilingualism has varied widely and, unfortunately, has not yielded the diversity results it has been expected to — not in terms of linguistic diversity and certainly not in equity at large. It is undeniable that English continues to be a powerful and useful lingua franca in the field (which has been noted in other disciplines, however it has also been noted that this creates an excluding barrier in academic discourse in the conference setting [Fregonese 2017]. However, it is also one that encodes modalities of writing and epistemologies of the academies that use it and implied notions of merit. Thus, among the most important obstacles is the human infrastructure required to evaluate proposals appropriately as we have argued in the section devoted to peer review. Likewise lacking is a conference infrastructure capable of accommodating multilingualism at every step of the conference process: from the conference management, to the publication of the CfP, to the reviewing of proposals, to the proceedings of live presentations, and then to the publication of conference papers.

The aspirations of the conference to be multilingual and for that multilingualism to be a conduit for equity and diversity are worth pursuing. Nevertheless, we believe that the prism of multilingualism that has been favored in ADHO, on its own, is not capacious enough to lead to more diversity, a just way of evaluating proposals, or a wider understanding of various academic cultures. Moreover, multilingualism is also too narrow to foster anti-hegemonic, anti-racist, anticolonial, gender-balanced conference practices and operations. Indeed, even in the instances where linguistic diversity was central to the conference, it alone has done little to address these parallel problems. Aside from instances where specific PC chairs have taken action, it has fallen upon groups outside of ADHO governance and of the various committees in charge of organizing the conference to bring these issues to the forefront. In recent years, Global Outlook::Digital Humanities (GO::DH), a special interest group, has advocated not just for incorporating multilingualism beyond the CfP itself and in presentations and other aspects of the conference but also underscored how equity and diversity are multiplex [GO::DH n.d.].[24] A more expansive and deeper understanding of equity of diversity that does not rely on language representation needs to be deeply and thoroughly integrated into conference organization and ADHO governance.

In consideration of these issues, we urge ADHO to designate funds to diversity efforts. This will involve rethinking the role, function, and ambit of the Multi-Lingual Multi-Cultural committee (MLMC). Multilingualism is not capacious enough for diversity and equity aspirations nor has it been evenly applied at past conferences. Similarly, we encourage Local Organizers to budget funds for equity and diversity measures from the beginning. Elements which should be considered in the budget include live and printed translation costs, accessibility measures (see below), and funding to support participants from underrepresented communities.

Lastly, we note that in order for ADHO to be accountable when it comes to its stated diversity commitments, the conference must gather data on diversity and equity and set clear timelines for goals, assessments, and consequences. ConfTool could be used to gather data related to biographical and demographic details. A standardized post-conference survey, which has not been utilized in previous years, could also be useful to gather information about what initiatives were effective or not in a given year. Further, initiatives demonstrated to be successful must have continuity, permanent financial support, and not be abandoned by future PC Chairs when they diverge from opinions held by chairs or their constituent organizations.

In any analysis of organizational behavior and processes, there is always a moment of exploring motivation. Throughout this paper, we’ve attempted to show how the best of intentions in the implementation of conference processes have, in fact, not led to the best of results. From our first-hand experiences and with conference data in hand, we offer the following suggestions to begin to improve the conference for conference organizers, participants, and beyond. While our suggestions may seem specific to ADHO, they may also serve as a gentle reminder to all conferences that their processes should be transparent and self-reflexive.