Digital Multilateralism in Practice: Extending Critical Policy Ethnography to Digital Negotiation Sites

The COVID-19 pandemic has significantly increased the use of online tools in the conduct of multilateral environmental negotiations. Although scholars have recognized that information and communication technologies have gradually been reshaping traditional diplomatic practice, such technologies are not considered to be transformative of diplomatic practice itself. However, owing to the COVID-19 pandemic states have had to rush into unprecedented and unpredictable forms of digital cooperation that are poorly understood. To illuminate this uncharted area, our research applies combined digital and critical policy ethnography to two online dialogues within the framework of ongoing negotiations under the United Nations Convention on the Law of the Sea toward a new treaty for the protection and sustainable use of marine biodiversity beyond national jurisdiction. Digital critical policy ethnography conducted at two online sites enables us to study the political effects of emerging international practices. We re-interpret digital diplomacy in terms of “communities of practice” developing across, and connecting physical and digital sites. Virtual communications amongst state and non-state actors mirror traditional forms of diplomacy whilst introducing new practices that may change conventional forms of international treaty-making. We propose the term digital multilateralism to capture these new forms and conclude that it can have two effects: deepen the background knowledge of actors that form a community of practice and create new inequalities.


Introduction
The COVID-19 pandemic has significantly increased the use of online tools for conducting multilateral environmental negotiations and reinvigorated the debate on the challenges and opportunities of digital diplomacy in the United Nations (UN) context ( Chasek 2021 ;. Given the extensive use of online meeting platforms by treaty secretariats and states seeking intergovernmental dialogue, a revived debate on digital diplomacy is needed. Scholars have acknowledged that information and communication technologies (ICTs) have gradually been reshaping traditional diplomatic practice-for instance, through increased immediacy ( Seib 2012 ) or by remodeling the relationship between state and citizens (e.g., Adesina 2017 )-yet ICTs are not viewed as transformative of diplomatic practice itself ( Hocking et al. 2012 ;Adler-Nissen and Drieschova 2019 ). However, the COVID-19 pandemic and the extensive use of new digital sites mirroring multilateral negotiation practice are challenging conventional views on digital diplomacy and the study thereof; new concepts and methods are needed.
Digital diplomacy has evolved into a "strategy of managing change through (…) virtual collaborations" ( Holmes 2015 , 15). Diplomats, foreign ministries, and embassies have considered digital diplomacy to be "significant" even though it is uncertain "how they should handle it" ( Hocking et al. 2012 , 48). The digitalization of diplomacy is thus becoming a key change factor for diplomatic practice in the twenty-first century, even though it seems to contradict key principles and enabling factors of international cooperation. Physical institutional settings enable trustful human-to-human interaction ( Coleman 2011 ), trust being a precondition for cooperation in multilateral negotiations ( Touval 1989 ;Chasek and Wagner 2016 ). Face-to-face meetings increase the likelihood of cooperation because they enable individuals to exchange information and empathize with each other, reducing uncertainty and increasing willingness to cooperate, even when they have reasons to distrust ( Holmes 2013 ). While in the past decades, these disabling factors of digital diplomacy dissuaded states from replacing physical settings with digital ones, the pandemic has thrust them into unprecedented and unpredictable forms of digital cooperation that are poorly understood ( Chasek 2021 ;. To illuminate this uncharted area, our research combines digital and critical policy ethnography in the ongoing UN negotiations concerning marine biodiversity. Arguing that online dialogues constitute negotiation sites despite their lack of legal validity, we conducted digital critical policy ethnography at two online sites: the Informal Intersessional High Seas Treaty Dialogue and the Virtual Intersessional Work of the Intergovernmental Conference . Digital ethnography enables the study of online dialogues from within ( Pink et al. 2016 ); critical policy ethnography, for its part, studies practices and processes relevant to policymaking, such as the formation of alliances and struggles over semantics across different sites and periods of time ( Shore and Wright 1997 ;Wright 2011 ;Wright and Reinhold 2011 ). Against this background, we asked whether digital diplomacy might replace in-person diplomatic practice, under which conditions and what effects this would have.
To conceptualize digital multilateral sites, we extend the notion of "field" to digital space and argue for the need to examine digital multilateralism's effect on policymaking. Practice approaches allow us to study online diplomacy in continuity with in-person diplomacy, that is to say, to make sense of changes in diplomatic interaction and view the (digital) negotiation site as a "field" where diplomats conduct diplomatic practices that are "empirically traceable" ( Pouliot and Cornut 2015 , 308). In addition to new digital diplomatic practices, whose political effects we studied through digital critical policy ethnography conducted at the two above-mentioned online dialogues, we are interested in processes of change and contestation embedded in emerging digital multilateral sites. We re-interpret digital diplomacy in terms of "communities of practice" emerging across physical and digital sites and connecting them. This enables the relational and comparative study of practices performed digitally and physically. A focus on processes of change and contestation is necessary to unpack the meaning of such practices, which may be highly relevant if digital multilateral negotiations become the common media of diplomacy in the years to come ( Chasek 2021 ).
In the following, we first explore how the literature addressed ICTs' effects on diplomacy before the COVID-19 pandemic. The expansion of existing notions to a multilateral setting forces us to go beyond digital diplomacy as a pre-pandemic practice and concept. In the second section, we define the key characteristics of digital multilateralism, show that critical policy ethnography can be extended to the study of international practices in digital space, and introduce the digital sites of the negotiations under study. Third, we explain how we applied digital critical ethnography to the two above-mentioned online dialogues, providing details on data collection and analysis. Fourth, we analyze digital diplomatic practices for each dialogue, including the ways in which they mirror traditional diplomacy or differ from it, and present our results. The fifth section discusses our results in relation to an emerging community of practice across physical and digital spaces and the ethnographic study thereof. We conclude that digital infrastructures not only reinforce inequalities amongst state and non-state actors but also create new ones while deepening the background knowledge of participant actors in online discussions about the Treaty and its provisions, other actors' interests, and overlap areas.

Beyond Pre-Pandemic Digital Diplomacy
The past two decades have seen a significant increase in "information-rich, highly interactive environments" ( Singh 2015 , 181) for diplomacy, to the extent that digital diplomacy is considered a key challenge of international cooperation in the twenty-first century ( Hocking et al. 2012 ). The "digital" in digital diplomacy implies a technology or a process that can be both online and offline ( Ardèvol and Lanzeni 2017 ). Although responding to new technologies has always been an integral part of diplomatic practice, the growth of ICTs challenges diplomats to adapt both practice and organizational capacity ( Hocking et al. 2012 , 48).
So far, practitioners and scholars alike have shared the view that changes induced by digital communication tend to be incremental, transforming diplomacy "but only at the margins and evolutionarily" ( Pouliot and Cornut 2015 , 307). This slow adaptation can partly be explained by the "old-fashioned tradition" of diplomatic practice, which favors coexisting with far-reaching innovation rather than changing its modus operandi ( Cohen 2013 , 30). For instance, incremental changes have occurred in bilateral diplomatic exchanges ( United States Institute of Peace 2007 ), public diplomacy activities (e.g., Adesina 2017 ;Bjola 2015 ;Collins, DeWitt, and LeFebvre 2019 ), and the interaction between consular representations abroad and citizens in the host country (e.g., Adesina 2017 ). In most cases, the temporal dimension of diplomatic relations has been examined: the acceleration of specific tasks, such as (consular) activities, information management, and policy planning ( Adesina 2017 ), and the immediacy of communications ( Seib 2012 ).
As an approach to managing change online ( Holmes 2015 ), digital diplomacy has been the object of research into the causes of resistance: Why does adapting to the digital world seem to be especially challenging in the diplomatic realm? A key factor for successful international cooperation according to the literature is the availability of physical institutional settings. These enable individual actors can exchange information and empathize with each other ( Coleman 2011 ), which reduces uncertainty regarding other actors' intentions and also reduces distrust ( Holmes 2013 ). In other words, face-to-face interaction is considered to significantly increase the likelihood of agreement between two or more parties; this is especially relevant in multilateral negotiations.
Thus, while diplomacy can adjust to digitalization-and indeed has, as demonstrated above-the replacement of face-to-face interaction with virtual interaction is especially difficult when it comes to developing trustful relations in multilateral negotiations, which are a simultaneous process involving at least three parties aiming for an acceptable outcome over one or more issues ( Touval 1989 , 159). The complexity of such talks has driven diplomats to develop highly structured, restrained settings ( Winham 1977 ), where practices differ from those performed in bilateral meetings ( Chasek 2001 , 24). These include the attainment of consensus and a decisionmaking rule specifying that abstention is an affirmative rather than a negative vote ( Zartman 1994 , 5), which makes a successful conclusion more likely but often results in toothless agreements. By their very nature, multilateral negotiations require "orchestrating," a role played by elected presidents, chairs, or facilitators supported by Treaty secretariats and other UN staff ( UNEP 2007 ).
Indeed, a number of authors acknowledge diplomacy's performative dimension as intersubjective and constitutive ( Constantinou 1994 , 23). There are stages, scripts, casts, and audiences ( Death 2011 , 8), while both process and results are influenced by location ( Salacuse and Rubin 1990 ;Henrikson 2005 ;Coleman 2011 ; Craggs and Mahony 2014 )-"Parties frequently negotiate long and hard about where they are to meet long before they (…) discuss what they will negotiate" ( Salacuse and Rubin 1990 , 5). As Craggs and Mahony (2014 , 420) argue, the designated time and space for a conference exclude actors that either were not invited or could not attend; they also interpret "other spaces and times as inappropriate for discussion" (2014,415).
Finally, the "international conference circuit in part reflects a Western-dominated diplomatic landscape" and symbolically contributes to maintaining diplomacy as a Western activity ( Craggs and Mahony 2014 , 420). Thus, the Western hemisphere occupies a higher position in the hierarchy at sites where international politics are played out such as conferences, which act as legitimation spaces ( Coleman 2011 ;Craggs and Mahony 2014 ) and are characterized by past diplomatic practices ( Henrikson 2005 , 370).
Digitalization and new ICTs, however, have also shaped multilateral conferences ( Coleman 2011 , 168). The use of word processing software and email, for instance, has speeded up negotiations in the Committee of Permanent Representatives of the European Union, heavily influencing the drafting process and facilitating compromises ( Adler-Nissen and Drieschova 2019 ). Suiseeya and Zanotti (2019) , who studied the United Nations Framework Convention on Climate Change Conferences of the Parties (COPs) consider digital spaces to be part of the diplomatic practice itself. By including the meeting's digital life-namely, the use of social media and online interfaces by observers-as part of the field site, they analyzed indigenous representation practices and emerging forms of contestation in the negotiation room.
While the combined analysis of digital and in-person practices is proving an essential avenue for future research, previous studies only constitute the first step. Due to the pandemic, states rushed into unprecedented and unpredictable forms of digital cooperation; these call for new approaches to study digital diplomacy and its role in the twenty-first century ( Chasek 2021 ;. Owing to the lack of precedence and predictability of online meetings in multilateral negotiations, practitioners and scholars alike are struggling to assess their meaning and relevance for policymaking and international cooperation in the future. We consider these emerging online dialogues as field sites of digital multilateralism. Drawing on insights from practice theory, critical policy ethnography, definitions of pre-pandemic digital diplomacy ( Singh 2015 ), and multilateral negotiation ( Touval 1989 ), we understand digital multilateralism as a set of digital and physical diplomatic practices performed across space and time by state and non-state actors engaged in a joint enterprise of simultaneous negotiation through physical and digital infrastructures in information-rich, highly interactive environments. By using the example of ongoing UN negotiations on marine biodiversity, we show how critical policy ethnography can be extended to digital negotiation sites, thereby enabling the empirical study of digital diplomatic practices in a multilateral setting.

Critical Policy Ethnography in Digital Settings
Ethnography of diplomacy focuses on social relations and the international practices that constitute cooperation and the making of the world order within specific settings for political action. For instance, Collaborative Event Ethnography-an innovative method designed "to study global environmental politics (…) in practice" ( Büscher 2014 , 132)-enables a team of ethnographers to observe policymaking practices on negotiation sites, as well as study the politics of performance ( Campbell et al. 2014 ;. Shared international practices, meanings, and agreement-making are at the heart of their work. In practice approaches to IR, international practices are "socially organised activities that pertain to world politics, broadly construed" ( Adler and Pouliot 2011 , 7) and constitute "socially meaningful patterns of action, which (…) simultaneously embody, act out, and possibly reify background knowledge" ( Adler and Pouliot 2011 , 4). They are performed by individuals and groups acting within the same interpretive setting, implying that the performance will be socially recognizable by those individuals and groups. As for the setting, it establishes "the terms of interaction (…) and provides the background knowledge of expectations, dispositions, skills, techniques, and rituals that are the basis for the constitution of practices" ( Adler 2019 , 110). Action in this regard produces repeated interactional patterns; it also seeks to impose meaning, either to uphold or contest the world order .

From In-Person to Digital Practices: Theory
Studying practices means engaging with the relationship between actors and their environment ( Adler and Pouliot 2011 , 2), and with how practitioners instantaneously change old practices in new situations ( Cornut 2018 , 721). To capture this relationship, some scholars draw on the notion of "communities of practice" ( Adler 2019 ;Wenger 1999 ), which are systems of linked participation ( Wenger 1999 ). They require the interrelated engagement of their members during a negotiation process that makes them accountable to each other because they use common resources repetitively ( Wenger 1999 ). A community of practice is not physically bounded and may change over time. It can perform in a digital setting because its members share background knowledge that transcends the digital sphere-about a specific context, the specific repetitive ways of acting and attributing meaning to specific action types, and framing the issue at hand. Thus, the boundary between the digital and the physical dissolves, implying that digital sites need to be studied in relation to physical ones by tracing diplomatic patterns of action across sites and time, comparing them, and interpreting their meaning.
In order to study this empirically, we propose to extend critical policy ethnography to digital sites. This approach investigates settings where actors perform practices to agree on new policies ( Dubois 2015 , 476). 1 It conceptualizes the field broadly as the political space for actors, practices, and institutions relevant for policymaking ( Shore and Wright 1997 ;Wedel et al. 2005 ;Wright 2011 ;Wright and Reinhold 2011 ). The field can expand or contract and usually operates on different sites that are not necessarily "geographically fixed" ( Wedel et al. 2005 , 39;Wright 2011 ;Wright and Reinhold 2011 ). Researchers follow policymaking at these various sites to analyze contestation processes across space and time as well as "interactions (and disjunctions) between different sites or levels in policy processes" ( Shore and Wright 1997 , 14-15;Wright 2011 ;Wright and Reinhold 2011 ).
While critical policy ethnography helps conceptualize online dialogues as policymaking sites, it is not well enough equipped to capture the specificities of the "digital." We complement it with insights from science and technology studies into digital communication and digital ethnography. Science and technology studies treat media technologies and the Internet as socio-material complexes ( Gillespie, Boczkowski, and Foot 2014 ), forms of "social shaping" ( MacKenzie and Wajcman 1990 ), and field sites in their own right ( Hine 2017 ). They acknowledge the multiple meanings that actors attribute to specific technologies and the numerous ways in which they are used ( Mol 2002 ). The context determines specific forms of use of digital space, which are constituted by "specific sets of practices through which digital spaces acquire meaning" ( Hine 2017 , 23).
Digital ethnography helps us find out how actors use technology in each context ( Hine 2017 , 23). A digital space where multilateral dialogue amongst state and non-state actors is performed thereby becomes an equally valid site for studying diplomatic practices. Moreover, emphasizing the impact of digital space on multilateral negotiations endows technologies with a "more precise role" in the practices of overall treaty-making ( Bueger and Gadinger 2015 , 453).

From In-Person to Digital Practices: Case Study
Since 2006, governments have been developing a new treaty to fill legal gaps in the United Nations Convention on the Law of the Sea (UNCLOS) related to ocean protection. It is intended to be a legally binding instrument for the conservation and sustainable use of marine biodiversity beyond national jurisdiction (hereafter BBNJ Treaty). The BBNJ Treaty may significantly change how we know and govern the ocean by regulating the following activities in the high seas: the utilization of marine genetic resources (MGRs); area-based management tools, including marine protected areas; environmental impact assessments; and capacity-building and transfer of marine technology ( Tessnow-von Wysocki and Vadrot 2020 ;. Key tensions have arisen among states since 2006, such as the conservation versus the use of marine biodiversity or the sharing versus the nonsharing of monetary benefits resulting from the exploitation of MGRs . While many Global South countries argue that the BBNJ Treaty must be governed under the "Common Heritage of Humankind" principle, implying that access to MGRs in the high seas and economic benefits associated with their utilization should be regulated, actors such as the USA, the EU, Japan, Russia, and Iceland uphold the "Freedom of the High Seas," especially in relation to ocean research . Governments have participated in thirteen prenegotiation sessions (2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017) ( Earth Negotiations Bulletin 2015 ;Earth Negotiations Bulletin 2017 ) and three Intergovernmental Conferences (IGCs) (September 2018, March/April 2019, and August 2019) at the UN Headquarters in New York ( Res. 72/249 ). These efforts resulted in a Treaty draft that governments were expected to finalize and adopt at the fourth and originally final IGC in April 2020.
However, the UN General Assembly (G.A.) postponed the IGC4 owing to the COVID-19 pandemic ( Dec. 74/543 ). Subsequently, two initiatives were established to enable the diplomatic exchange of views between governments using digital communication platforms : the Informal Intersessional High Seas Treaty Dialogue and the Virtual Intersessional Work of the Intergovernmental Conference .
The Informal Intersessional High Seas Treaty Dialogue (hereafter "High Seas Dialogue") was initiated in April 2020 by Belgium, Costa Rica, and Monaco together with the High Seas Alliance (a group of NGOs). It uses a video conferencing platform, Webex, which creates virtual rooms that participants access by clicking on a link sent by organizers one day before the dialogue starts. Participants identify themselves with their name (in many cases, delegates add the state or NGO that they represent) before "entering" the virtual room. Ten minutes after the start, the room is locked and participants who have not identified themselves are shut out. The site is composed of small windows that display attendees' virtual presence and a panel with access to the chat function and the participants' list.
The Virtual Intersessional Work of the Intergovernmental Conference (hereafter "BBNJ Intersessional Work") was launched by the President of the negotiations in September 2020 to facilitate the written exchange of opinions amongst delegates through the chat functionality of MS Teams. Participants access the platform by logging into their private MS Teams account. Only registered participants may access BBNJ Intersessional Work by clicking on the United Nations Group. Registration conditions are the same as for any other UN negotiation. Actors registered for the formal negotiation in New York were asked for a new accreditation letter, to be sent either by their government, IGO, or NGO. On entering the site, we see a large dialogue window, where participants see others' posts. They can upload/download documents to/from the forum, and have private or group conversations with other participants. Since March 2021, BBNJ Intersessional Work facilitates verbal communication and exclusively uses Webex.

Material and Methods
At a digital multilateral "event," ethnographers observe and record policymaking practices and the politics of performance using empirical material collected on the digital site through mediation ( Beaulieu 2017 , 33). We applied digital critical ethnography to diplomatic practice in digital space by exploring specific patterns of action. The restrained setting of multilateral diplomacy has implications for "gaining access" that goes beyond the ability "to capture interactions or behaviors of interest" ( Beaulieu 2017 , 34). However, Vadrot (2020) indicates that such a setting includes accreditations to closed sessions and restrained researcher behavior-similar to ethnographic work-in the physical negotiation room.
We designed our ethnographic approach specifically for the two above-mentioned UN online dialogues: the "High Seas Dialogue" and the "BBNJ Intersessional Work." We followed an iterative procedure and used the same data collection method that we had used before the pandemic at the "physical" site. This enabled us to compare the two digital dialogues with our in-person observations in New York.
In total, 24 sessions of the High Seas Dialogue and 12 rounds of BBNJ Intersessional Work were observed between June 2020 and December 2021. In the case of the High Seas Dialogue, the study's Principal Investigator obtained access after contacting the organizers and could participate in Webex meetings via a personal link to the virtual room upon disclosure of identity. In the second case, all team members accredited to the BBNJ Treaty negotiations as part of the International Studies Associations' delegation received invitations to BBNJ Intersessional Work.
A matrix designed for the collection of field data in the physical negotiation room was applied to both digital sites. Five members of the research team collected and maintained the data; each session was observed by at least two team members. Regular debriefing about our observations allowed us to develop a shared understanding of the sites, their specificities, dynamics, and emerging or changing patterns of interaction amongst online dialogue participants.
The data consists of field notes added to the matrix, including entries with our observations of the site, (order of) speakers, (type of) statement, references to Treaty draft provisions, and time of statement. We collected background documents (e.g., participants list, summaries) circulated to participants, documents shared by delegations on the online platforms, chat histories, and screenshots for documenting general procedures and recording important moments and developments throughout the dialogues.
We analyzed the data by identifying socially meaningful patterns of action (digital practices), and comparing the digital sites (e.g., participants, rules, or issues) and practices (e.g., formation of alliances, absence, or appeasement). We also compared the results of our virtual data analysis with the results of earlier, in-person data collection in New York in order to identify elements of digital multilateral diplomacy and a community of practice, as well as processes of contestation and change across time and space.
We only considered data that were relevant to unfolding patterns of actions and what they might reveal about digital multilateral diplomacy. In accordance with the Chatham House Rules that apply to the High Seas Dialogue and the unofficial status of the BBNJ Intersessional Work in overall treaty-making, we refrained from analyzing the behavior of single state or non-state actors, their statements and positions, or the degrees of similarity or difference amongst state preferences. We also refrained from displaying information such as names, pictures of individuals, and screenshots.

Results: Emerging Practices On Digital Multilateral Sites-Insights from Two Virtual Dialogues
Our digital critical ethnography of the High Seas Dialogue and BBNJ Intersessional Work shows that both settings are built around digital infrastructures and modalities mimicking multilateral in-person interaction within the negotiation room at UN Headquarters (see Table 1 ), while having their own dynamics. Practitioners elaborate on previous practices in new situations, automatically changing these practices ( Cornut 2018 , 721), and, as Pouliot and Cornut (2015 , 306) argue, new practices evolve from past ones "in order to resonate in the present." The data also reveal significant differences between the two online dialogues at the time when these took place on two separate platforms, as described in Table 1 . The High Seas Dialogue was initiated in spring 2020 to keep up the BBNJ momentum and provide both state and non-state actors with a virtual videoconference room for exchanging views on the Treaty draft.
BBNJ Intersessional Work started later (September 2020) by the President of the negotiations in response to the lack of a UN-driven online dialogue and a formalized framework for the online exchange of views on Treaty draft. Via MS Teams, accredited delegates (one person per delegation and treaty package element) were allowed to post comments and upload positions for five consecutive days. Since March 2021, BBNJ Intersessional Work has also held several 3-hour sessions using Webex that were attended by more than 300 participants from different countries and organizations.
Although it might not be the case for every single delegation, we observed a considerable overlap of delegates, both state and non-state actors, across the two online dialogues. In the case of the High Seas Dialogue, the continuous engagement of delegates has resulted in the verbal expression of a shared understanding of the use of this digital tool and its purpose. Indeed, in sessions in December 2020 and February 2021, chairs and facilitators highlighted that "there is a growing level of comfort [within these dialogues] as we see each other monthly" 2 and that "there has been an evolution of the positions." 3

Digital Diplomatic Practices
For groups of participants to be labeled "communities of practice," they need to engage in a process using a shared repertoire in a repetitive way ( Wenger 1999 ). This implies that (new) digital practices stem from existing practices, i.e., those we observed in person at UN Headquarters. In the following, we describe digital diplomatic practices and indicate where we detected continuity and where we found change compared with traditional diplomatic practices. We also describe processes of contestation and change.

High Seas Dialogue
The Webex platform has enabled specific diplomatic practices that are synchronous with the past, such as the presence of "orchestrators," the opportunity to join "coffee talks," and the use of rules and procedures. In each Table 1. Comparison of the two online dialogues and the in-person BBNJ Treaty negotiations by using the criteria for multilateral negotiations in Chasek (2001)  observed session, the "orchestrators"-the High Seas Alliance organisers and one state representative (from either Monaco, Costa Rica, or Belgium)-would introduce the dialogue's modalities and purpose. They would introduce participants to the rules and procedures (Chatham House rules) and explain that this was a space of trust where Treaty draft could be openly discussed, not a formal negotiation setting-a practice designed to appease the proceedings and keep states on board. Then they would introduce the two facilitators (UNCLOS experts) in charge of orchestrating dialogue amongst participants and ask state and non-state actors to use the chat function to signal whenever they wished to make a statement.
To provide the space to meet informally, the organizers encouraged delegates to join the dialogue 5 minutes early for attending "coffee meetings" as if they were in the UN Headquarters. Talks over coffee are common among delegates in in-person negotiations and are often used to bargain and build trust.
Similarly to the formal in-person procedure, both state and non-state actors had to "ask for the floor" in order to "speak," while chairs and facilitators would "give the floor to" a specific actor, and, in return, state and non-state actors formally thanked chairs and facilitators. Statements made by state representatives have tended to mirror the drama of negotiations: For instance, strong emotional statements indicating a state's position and its red lines, or delegates refraining from answering a question asked by the chair if it addressed an issue that had not yet been discussed within the delegation or the alliance.
When discussing principles applicable to the BBNJ Treaty, Global South states backed the "Common Heritage of Humankind," while Global North states opposed it by calling for the "Freedom of the High Seas"; this led to tension rising during the online dialogue. The prevalence of the "Freedom of the High Seas" in the BBNJ Treaty would continue enabling the privatization of MGRs found in international waters.
Statements about the status of non-parties to UNCLOS also mirrored the drama of negotiations. Non-parties emotionally and explicitly indicated that they would not sign the BBNJ Treaty if this changed their status. Otherwise, UNC-LOS parties could argue that, by signing an implementing agreement of UNCLOS like the BBNJ Treaty, non-parties would automatically recognize UNCLOS' provisions. They also requested a non-adversarial implementation and compliance mechanism to avoid court disputes between parties and non-parties to UNCLOS, as the settlement of such disputes is resource-intensive.
The formation of alliances, which is a central feature of multilateralism, has persisted on the virtual platform: One delegate speaks on behalf of all European Union Member States, while alliances such as the Core of Latin American Countries, the Pacific Small Island Developing States (PSIDS), or the African Group also speak with one voice-showing that they have continued exchanging views amongst themselves and developed a common position despite the pandemic. Furthermore, state actors align themselves with other states by supporting particular statements, indicating that diplomats build on in-person multilateral diplomatic practice to interact in digital space.
In one of the online dialogues of February 2021, at the end, the President of the negotiations took the "floor," encouraged the continuation of online discussions and highlighted their importance for overall treaty-making. Closing words by presidents and chairs are very common and are often used to express gratitude toward governments for engag-ing in negotiations and to encourage further cooperation. In this case, it was especially remarkable because it increased the legitimacy of an online dialogue that had repeatedly been described as "non-negotiation" and constituted it into an important element for a successful conclusion of the BBNJ Treaty. In line with the email sent by the organizers, the subsequent High Seas Dialogue purposely took place after the BBNJ Intersessional Work webinar, as discussed with the President of the negotiations. This further increased the High Seas Dialogue's legitimacy in the treaty-making and shored up diplomatic practices performed in this setting.
We have also observed practices that differ from those performed on-site. For instance, facilitators asked delegates to clarify their positions, encouraging them to elaborate on their statements and, sometimes (when issues had not been discussed within the delegation), to "change their hat" and respond from a personal point of view. This practice of changing hats may also be observed when other state actors ask a delegate to clarify their state's position. This enhanced dialogue between state delegates and the more dynamic nature of digital interaction has sometimes increased the clarity of the kind of Treaty and Treaty provisions that states are striving for. It has enabled participants-and us as observers-to dive much deeper into technical details. The impression that digital dialogues are "more qualified" than formal negotiations resonates with the fact that during the chat, participants tend to post links to scientific papers or issues while scientists and other experts post short contributions.
Small delegations, however, have not been able to participate in all High Seas Dialogue sessions. This has hindered their chances of diving much deeper into technical details, obtaining explanations from scientists and other experts, and quickly accessing scientific papers distributed on the digital site. This disadvantageous position carries with it the risk of enabling larger delegations-or wealthier countries-to dominate the digital setting.
Although it is not a small coalition, the Group of 77 and China have not actively participated in the High Seas Dialogue, but several of its members did attend online meetings. Its lack of engagement might be due to the difficulty of solving internal differences; its members cannot all meet digitally at the same time because they are scattered across many time zones.
Actors taking the floor were asked to turn their cameras on unless they had a weak internet connection, which was most often the case with Global South participants. We noticed that Global North actors had their cameras switched on more often. Thus, even though Global South participants interacted in the discussions, they could not fully communicate with other delegates. This also favored Global North actors, who could more frequently communicate non-verbally in the digital setting.
Only the English language has been used, and statements cannot be issued in any other language because no translation services are available. This excludes delegates who are not fluent in English (i.e., African Group's delegates) and gives English-speaking delegates an advantage during digital discussions.
By taking the floor when no state actor asked for it and there were moments of silence, non-state actors have driven the dialogue forward. This unusual situation has given non-state actors the opportunity to influence the discussions differently than during in-person negotiations: They encouraged states to continue exchanging positions.
The role and nature of the High Seas Dialogue have changed over time. It started as a virtual room systematically discussing the four package elements of the Treaty, guided by specific questions introduced by facilitators and circulated to participants beforehand. After the start of BBNJ Intersessional Work, the High Seas Dialogue started to imitate it: It gave participants the opportunity to elaborate or reflect on statements made in BBNJ Intersessional Work, a dynamic that connects the two online dialogues and their policymaking processes.

BBNJ Intersessional Work On MS Teams
Digital practices observed during BBNJ Intersessional Work were partly similar to those observed during the High Seas Dialogue. The President of the negotiations explained in an opening statement that online work was no formal negotiation; whether results might lead to a new draft before the final IGC was kept open. This practice of appeasement allowed states holding divergent views on the status of online work to participate.
Further observations indicate that delegates' digital practices resemble those performed during in-person negotiations: the use of polite language, the sharing of documents, and the formation of alliances. Rituals appropriate in a diplomatic setting have been maintained on MS Teams, such as thanking the chair at the beginning of a contribution or starting written statements with very formal acknowledgments.
The formation of alliances has served to structure interaction amongst participants, including referring to and supporting other states' positions. State and non-state actors explicitly mention in their forum statements whether they agree or disagree with each other. For instance, Pacific states agreed to have an enabling clause in the BBNJ Treaty that would permit the COP to commission strategic environmental assessments and clarify their purpose. 11 Other states converged on the need for this enabling clause, a step forward in overall treaty-making.
State and non-state actors have also engaged in contestation practices. The divide between the Global North and South persisted with regard to the principle that should apply to MGRs: "Common Heritage of Humankind" or "Freedom of the High Seas." In other words, Global South governments contested the current world order by trying to decrease the uneven exploitation of marine resources, while Global North states continued protecting their economic interests in ocean space .
Both state and non-state actors can also upload documents (positions, scientific articles, and other publications) to MS Teams (similar to papersmart 12 ) to hand their statements in. For instance, the UK uploaded documents every time it posted an answer. PSIDS uploaded documents after posting some answers. A wider range of NGOs-compared to states-has uploaded scientific or institutional publications (e.g., IOC's non-paper on existing and potential contributions of IOC-UNESCO to the BBNJ process).
Delegates also performed new practices, such as the passive participation of otherwise active state actors, a limited number of "speakers" per delegation, and delayed answers by state actors. Although a registered member of the Intersessional Work, the Russian Federation, for instance, has refrained from making statements online. This may 11 Strategic environmental assessments are evaluations of the potential environmental and health effects of an activity (revised draft text of an agreement under UNCLOS for the conservation and sustainable use of BBNJ, art. 1.13).
12 Papersmart was a platform where UN meetings' participants uploaded their written statements and accessed other participants' statements.
be interpreted as a means to delegitimize digital space for discussion purposes and introduces a new meaning to the role of abstention. More specifically, abstaining from digital participation implies disagreement with conducting online discussions and using their results in the following IGCs. Only accredited persons may access the digital site, but not every accredited person is allowed to post statements. Each delegation (state or non-state) was asked to select one person per package element and only this person may produce a written statement on behalf of their delegation or alliance. Given that each session runs for five consecutive days-making it possible for actors to connect to the system anytime-states do not have to answer right away when a question has been addressed to them, which makes the session less dynamic and immediate. The aim is to allow states to have a more comprehensive exchange of views and positions and give small delegations the same opportunity as large ones to engage in online communication.
Just as in the physical negotiation setting, NGOs may only "speak" after states. In practice, however, this is not always the case because state actors can take many days to post an answer and facilitators do not indicate from which point in time non-state actors are allowed to speak. In other words, there is no way to control the order in which participants express their views. This implies that the influence potential of non-state actors in online discussions lies in their posting comments ahead of state actors and expecting that state actors will use the provided information in their subsequent comments.
Although delegates may post answers in any official UN language, the lack of translation services makes English the de facto language of online discussions. The African Group has pointed to the difficulties of many of its delegates to participate in the forum because of the lack of translation. Moreover, African Group delegates uploaded documents in French. As with the High Seas Dialogue, the overriding use of English has benefitted English-speaking delegates and excluded those who are not fluent in this language.

BBNJ Intersessional Work On Webex
BBNJ Intersessional Work conducted on Webex features similar practices to those found during in-person negotiations and the digital High Seas Dialogue. For instance, delegates have continued to use very politely, diplomatic language and to express their thanks for the opportunity to take the floor. They also post links to online documents (such as policy briefs) on the chat function but refrain from uploading documents to the digital platform. Alliances have also been formed and states openly agree or disagree with each other's statements. For instance, non-parties to UNC-LOS explicitly agreed to not give the COP the opportunity to seek advisory opinions in tribunals when compliance issues arise.
The online dialogue on Webex resembles the drama of both in-person negotiations and the digital High Seas Dialogue, with diplomats restating positions that they are not willing to modify. States from the Global North and Global South continued advocating different principles in relation to MGRs, and non-parties to UNCLOS continued fighting to protect their status in the BBNJ Treaty and obtain a non-punitive implementation and compliance mechanism.
Similarly to the High Seas Dialogue, English is the de facto language, reinforcing the exclusion of delegates who are not fluent in this language. Also, Global South participants more often had their cameras turned off in comparison to Global North delegates, hindering their non-verbal communication and disadvantaging them during online talks. Finally, small delegations have not been able to attend all sessions, leaving room for large delegations to dominate discussions.
Common practices across the High Seas Dialogue and BBNJ Intersessional Work include the use of diplomatic rituals; the formation of alliances; the sharing of documents; the use of rules and procedures; and the use of the English language. Unlike in BBNJ Intersessional Work on MS Teams, delegates gave emotional statements that conveyed their red lines in the dynamic discussions of the High Seas Dialogue and BBNJ Intersessional Work on Webex. Responding to the lack of spaces for highly informal conversations, the High Seas Dialogue provided "coffee talks" before discussions; BBNJ Intersessional Work, for its part, failed to provide informal spaces.

Discussion: the Future (Study) of Digital Multilateralism
Prior studies emphasizing the importance of studying ICTs' effects on diplomacy have demonstrated that new information technologies have been transforming diplomacy in line with established practices ( Cohen 2013 ;Pouliot and Cornut 2015 ;Singh 2015 ). This study set out to problematize unprecedented digital multilateral sites that emerged during the COVID-19 pandemic and the study thereof. We started from the observation that the UN employs ICTs to create digital spaces, allowing interaction amongst governments. This has been especially important for the negotiation of new treaties, as exemplified by the BBNJ negotiations. The Treaty was to be concluded in April 2020 after four IGCs and faced the risk that governments might lose interest during the pandemic. Against this background, we asked whether digital diplomacy might replace in-person diplomatic practice, under which conditions this might take place, and what effects it would have.

A Community of Practice Across Digital and Physical
Sites?
Our data suggest that digital diplomacy conducted on the two platforms contains both elements of continuity and change with regard to traditional (in-person) diplomatic practice. Practitioners instantaneously change old practices when facing new situations ( Cornut 2018 , 721) and innovative practices only resonate in the present if they relate to the past ( Pouliot and Cornut 2015 , 306). Many delegates are happy to continue negotiating their "joint enterprise"-the BBNJ Treaty-by adapting to innovation and employing a shared repertoire ( Wenger 1999 ) of international practices (formation of alliances, absence, and appeasement). Previous work shows that most participants gained access to both online dialogues because "they were registered for IGC4" and many had engaged in the BBNJ process since IGC1 ( Vadrot et al. 2021 , 181). We also observed a significant amount of overlap and synchronicity between digital and physical sites as regards participants and the rules and procedures they follow. Although the digital infrastructures used cannot fully replicate the physical negotiation room, patterns of action share important elements and follow essential rules that are unique to multilateral settings while introducing new practices. Our data also show evidence of cross-fertilization between the two online dialogues, awareness amongst participants of developments in each of them, and symbolic action (e.g., opening and closing statements by the President) elevating the status of online work to usual treaty-making.
The data show that delegates had previously created a trust basis or they had not continued cooperating in the online dialogues ( Touval 1989 ;Chasek and Wagner 2016 ). BBNJ Intersessional Work took this trust basis for granted by failing to openly provide sideline spaces for informal conversations, while the High Seas Dialogue encouraged cooperation through "coffee talks." Sideline or coffee conversations fulfill delegates' need to meet informally for building trust and resolving issues, but this can hardly be met on digital sites ( Coleman 2011 ;Holmes 2013 ;Chasek 2021 ). For instance, delegates had to communicate with all attendees of the High Seas Dialogue coffee talks and could not openly have a conversation with a particular delegate.
The results show that specific contestation processes traveled into and across digital sites. In all pre-negotiation sessions and IGCs, states struggled over the principle that should apply to MGRs . Such struggles persisted in the High Seas Dialogue and BBNJ Intersessional Work, indicating that parties will hardly reach an acceptable outcome regarding the MGRs issue at the IGC4. Moreover, in both online dialogues non-parties to UNCLOS fought to keep their status should they become a Party to the BBNJ Treaty. It can be assumed that this struggle will also take place at the IGC4. Delegates clearly deemed the online dialogues an appropriate space to perform contestation practices.
The data also show that if state actors reach convergence on an issue, this issue is not discussed any further on other digital sites. This is exemplified by the enabling clause for the COP with regard to strategic environmental assessments. During BBNJ Intersessional Work on MS Teams, state actors converged on the need to include such a clause in the Treaty. Subsequently, this issue was not dealt with during the High Seas Dialogue. This signifies that digital communication can move the dialogue forward and that digitally interacting state actors focus more on contentious issues.
Many delegates, however, do not engage in digital space because of different time zones (i.e., Pacific states), language limitations (i.e., African states), or the unclear status of digital discussions within the overall treaty-making (i.e., Russian Federation). Building upon Wenger (1999) , only those delegates who actively engage in online dialogueseven with the camera turned off-constitute a community of practice and can develop a better understanding of each other's positions. As a result, all other actors lack the insights and technical information gained through digital discussion, which is bound to restrict their background knowledge and weaken their position during in-person negotiations. This is also the case for new diplomats. While they will learn about diplomatic practices by continually engaging in online dialogues, their background knowledge will be more limited than that of those who have physically attended the negotiations. Face-to-face communication is essential in international negotiations ( Salacuse and Rubin 1990 ;Holmes 2013 ).
Following Adler (2019) and Wenger (1999) , we argue that BBNJ delegates who have attended the online dialogues constitute a community of practice because; firstly, they have adapted to changes introduced by specific digital tools-Webex and MS Teams-in order to continue exchanging information despite the pandemic; secondly, they assess the appropriateness of their actions in digital space by drawing on traditional diplomatic practices; and thirdly, in digital space, they have continued both disputing certain issues and reaching convergence on others. When facilitators declare that BBNJ delegates are used to interacting through online tools because they meet monthly and that positions have evolved through the continuous use of these tools, they discursively reify the BBNJ community of practice ( Wenger 1999 ). This acknowledges sustained relationships within the community and its rapid adaptation to innovative tools ( Wenger 1999 ), as well as trust amongst members ( Touval 1989 ;Chasek and Wagner 2016 ).
Particular limitations weaken our assumptions about emerging forms of digital multilateralism: The criteria of simultaneousness, immediacy, and inclusiveness cannot be met at the same time and within a single digital infrastructure. Delegations involved in large international negotiations, such as the COP15 of the Convention on Biological Diversity, are still waiting to meet in-person to continue their work ( CBD 2021 ). In the case of BBNJ Intersessional Work on MS Teams, the president, chairs, and facilitators could not control the speaking order of state and non-state actors, and non-state actors' influence on discussions was reduced to posting comments ahead of state actors, since they could not meet delegates physically in the negotiation room ( Chasek 2021 ;. As for the High Seas Dialogue and BBNJ Intersessional Work on Webex, they are scheduled in the afternoon, Central European Time, but many actors live in other time zones, challenging inclusiveness and participation. This points to the issue of whose time zone is preferred ( Henrikson 2005 ). By setting a time for discussions that fits the working hours of diplomats in Europe and the Americas, the High Seas Dialogue and BBNJ Intersessional Work on Webex have put some states, most notably Pacific states, at a disadvantage, and have symbolically contributed to keeping the Western hemisphere on the higher rungs of the hierarchy in diplomatic space ( Craggs and Mahony 2014 ).
During in-person negotiations, the dynamic unfolds under time pressure and states have only limited speaking time. In contrast, in MS Teams, this dynamic is disrupted by the lack of simultaneous interaction. The dramaturgy is restricted to the organization of space and time, lessening the importance of theatre ( Death 2011 ) and emotion ( Wong 2016 ). Performance is limited to drafting the written comment (i.e., choosing words, using bold letters, or referring to scientific articles), providing space for state and non-state actors to negotiate on the basis of arguments that are detached from emotion.
While negotiation practices related to digital and physical multilateral spaces overlap considerably, abstention is a clear example of how actors may endow a practice with new meaning in digital space. Abstention plays a positive role at physical sites ( Zartman 1994 ), but our findings show that it plays a negative role at digital ones. The digital practice of abstention emerges from existing behavior at physical sites. It is endowed with the meaning of "disapproval," arising from the fact that current rules and procedures fail to recognize online dialogue as formal treaty-making. The performance of abstention on digital platforms might carry its meaning over into future in-person negotiations, delegitimizing the use (at physical sites) of digitally exchanged information and either intensifying or introducing new tensions during in-person negotiations.
Examples of digital abstention include the Russian Federation, which neither contributed to the High Seas Dialogue nor to BBNJ Intersessional Work, but was a very active actor on-site in New York. Another example is Iceland, performing different physical and online practices that nonetheless all shared the same objective: securing a weak BBNJ Treaty.
During in-person negotiations, Iceland sought to safeguard its current fishing quota even if this negatively affected marine biodiversity. During its appearances at the High Seas Dialogue and BBNJ Intersessional Work on Webex, Iceland strongly-and emotionally-defended its position, namely, maintaining the status quo regarding fisheries. However, it did not participate in BBNJ Intersessional Work on MS Team, which may be interpreted as disapproval of nonsimultaneous online discussions. By actively performing diplomatic practices designed to secure a weak BBNJ Treaty, both during in-person negotiations and online dialogues on Webex, Iceland embodies continuity. At the same time, it has changed one of its diplomatic practices by abstaining from BBNJ Intersessional Work on MS Teams to protect its interests.
A possible explanation for this might be that abstaining states seek to maintain the legal status quo and control ICTs' use in multilateral negotiations. Legally speaking, current UN online dialogues cannot be considered formal treaty-making stages. This means that states not participating in online work can block any use of its outcomes at the next stage of the formal negotiations and delegitimize the whole process. However, at the same time, they are excluding themselves from the BBNJ community of practice and damaging their chances to influence the (digital) policymaking process.
Another possible explanation for the practices performed by Iceland might be that, while BBNJ Intersessional Work on Webex only allowed for written communication, the High Seas Dialogue has provided an opportunity to continue exchanging positions in a similar fashion to in-person diplomacy, which enables delegates to express positions and emotions that contribute to mutual understanding and to communicate in a non-verbal way ( Salacuse and Rubin 1990 ;Holmes 2013 ;Wong 2016 ).
Equally, it might be interpreted as a protest against the use of (digital) locations for online intersessional discussions that were not negotiated by the parties beforehand ( Salacuse and Rubin 1990 ;Craggs and Mahony 2014 ). Moreover, it might signify disapproval of the expansion of online tools for conducting multilateral environmental diplomacy ( Craggs and Mahony 2014 ). It can therefore be assumed that digital multilateral sites, alongside the development of suitable digital infrastructures (e.g., dealing with inclusiveness) and rules of procedures, will become a contested issue in world politics.

The Future of Digital Multilateralism and the Ethnographic Study Thereof
One primary difficulty with the study of the two BBNJ online dialogues is that they fail to qualify as legitimate settings for negotiation ( Coleman 2011 ) and formal stages in treaty-making with official validity. The President of the negotiations and the Secretariat may not use online government statements to develop a new Treaty draft as a basis for the following negotiation rounds. However, the fact that diplomats have been meeting on digital platforms to advance Treaty-related discussions does turn the two online dialogues into relevant political sites for BBNJ treaty-making. By focusing on policymaking processes and widening the conceptualization of the field, we were able to apprehend the two online dialogues presented above as new sites for BBNJ negotiations. Conducting digital critical ethnography, we followed treaty-making to these new sites in order to trace continuity and change with regard to traditional Table 2. Comparison between on-site and digital ethnography of the BBNJ negotiations by using criteria for digital ( Pink et al. 2016 ;Beaulieu 2017 ;Hine 2017 ) and critical policy ethnography ( Shore and Wright 1997 ;Wright 2011 ;Wright and Reinhold 2011 )  Thanks to on-site observation, we gained a more comprehensive understanding of the treaty negotiations. In this regard, tracing continuities allowed us to understand online discussions, while tracing changes helped us understand practices' meaning in future in-person negotiations, as well as new ways of integrating digital diplomacy into multilateral environmental agreement-making. Changes introduced by digital spaces might intensify existing tensions or create new ones during in-person negotiations. For instance, some delegates might use information exchanged online to strengthen their positions or create new bargaining space, while other delegates might strongly oppose the use of such information to make deals.
Although online dialogues seem to be more inclusive arrangements for intersessional work because all delegates are invited to participate, in fact, they provide a political space where only a small group of states can reach an understanding, achieve convergence, and agree on compromises that might have an effect during in-person negotiations, thus intensifying existing inequalities or creating new ones in overall treaty-making.
Both of the observed online dialogues disregarded considerable time differences between world regions, thereby excluding Pacific state and non-state actors, and favoring the Western hemisphere. Yet in principle, digital discussions could be carried out at different times in order to accommodate several time zones and could provide translation services ( Chasek 2021 ). These changes would enable more equitable discussions in digital spaces that do not automatically reproduce the current world order.
Following BBNJ treaty-making to two digital sites enabled us to document similarities and differences between the traditional and digital ethnographies of agreement-making sites, as well as between the types of data that can be collected. Table 2 summarizes this information.
Digital critical ethnography accounts for treaty-making practices on new digital sites, but it has limitations. First, digital ethnographers cannot become part of the online community they research due to a lack of personal contact. Researchers are unperceived on digital sites as they are not allowed to make statements, meaning that they do not perform digital practices and limiting their background knowledge. Second, digital critical ethnography cannot observe delegates' WhatsApp groups or private chats on digital platforms because of access restrictions. Third, it cannot explain delegates' subjective points of view. Interviews with individual diplomats might solve this issue, yet we consider that conducting interviews during or after the coming IGC will provide richer data with regard to delegates' experiences both on digital and physical sites.
To develop a full picture of the political effects of digital multilateral sites, additional studies will be needed, diving deeper into the technical aspects of the digital infrastructures used and the significance of digital space in multilateral diplomacy ( MacKenzie and Wajcman 1990 ; Gillespie, Boczkowski, and Foot 2014 ;Hine 2017 ). These studies will be essential to understand digital practices because infrastructures and locations "enable certain types of behaviour" ( Bueger 2014 , 402). BBNJ Intersessional Work on MS Teams, for instance, opens prospects for scholars to focus on how ICTs shape written diplomatic communication ( Adler-Nissen and Drieschova 2019 ), while the High Seas Dialogue and BBNJ Intersessional Work on Webex enable research to explore the (un)effective communication of emotions through a digital interface and the disadvantages enforced by temporal disparities . What are the boundaries of digital platforms? What are the (political) geographies of treaty-making? What digital infrastructures are needed to support multilateral negotiations? What are the implications for participation and inclusiveness?

Conclusion
This article challenged pre-pandemic digital diplomacy and proposed a new approach suitable for a multilateral context. We conceptualized digital multilateralism as a set of digital and physical diplomatic practices performed across space and time by state and non-state actors engaged in a joint enterprise of simultaneous negotiation through physical and digital infrastructures in information-rich, highly interactive environments.
Digital diplomatic practices in a multilateral context carry meaning insofar as they enable delegates to dispute certain issues, reach convergence on others, create new inequalities, and intensify existing ones. Scholarly literature on diplomatic practice and digital diplomacy is bounded to unpack the meaning of new digital practices in each constellation and highlight the challenges faced by state and non-state actors in digital multilateral sites, such as equity and inclusiveness.
Online dialogues constitute negotiation sites in the policymaking field of marine biodiversity. They reduce delegates' need to travel to attend intersessional meetings personally and the carbon footprint of in-person conferences ( Chasek 2021 ), but how can we ensure that digital multilateral sites enable delegates to build trust in the (online) process? Issues of leadership, inclusivity, transparency, and security need to be addressed-and the notion of simultaneousness (re)considered-while developing digital infrastructures. Digital space does offer new opportunities for multilateral environmental negotiation. The BBNJ Intersessional Work and High Seas Dialogue might influence emerging forms of regional, digital intersessional work where time zone differences do not play a role ( Henrikson 2005 ). Digital infrastructures could then meet the criteria of simultaneousness, immediacy, and inclusiveness. While delegates might refrain from making final decisions online-as the BBNJ community of practice-a digitally sustained intersessional exchange of views that is fair and equitable would increase the online dialogue's legitimacy while enabling scientists and other experts to help delegates reach evidence-based decisions.
Hybrid models-which combine physical and digital arrangements-might allow for the continuation of multilateral meetings ( Chasek 2021 ). The BBNJ negotiations' case shows that equity and fairness must be addressed in order to develop deeper, more comprehensive treaties that will protect life on Earth.

Funding
This work was supported by the European Research Council (ERC), Grant No 804,599, Grant holder Alice Vadrot. Since October 2021, Silvia Ruiz Rodríguez is funded by the sowi: docs Fellowship Programme of the University of Vienna.