Governing with conversation culture – conditioning organizational interaction in a digital social movement

ABSTRACT

Digital platforms support forms of collective action with varying degrees of organizationality – that is, with organization-like attributes such as identity, actorhood, and decision-making. These forms include seemingly non-organized online groups, which can be easily dismissed as informal chatter. Building on communicative theories of organizing and Luhmann’s Theory of Social Systems, we explore the communicative organizationality of a Facebook group related to urban planning. By analyzing private chat logs of the group administrators, we explore the group’s organizationality and the ways in which organizational and interactional communication become intertwined to produce and curate digital political discussions. Our findings illustrate how technological affordances built for moderation also support the strategic organization of discussions to orchestrate political talk. We conceptualize such operations as discursive conditioning of organizational interaction and suggest that political action online can be propelled not only by connecting people, but also by the powerful yet concealed tactic of conditioning public online discussions.

KEYWORDS:

• Social movements
• digital activism
• organizationality
• CCO
• theory of social systems

Introduction

Online environments offer new ways to organize political participation and collective and connective action (Bennett & Segerberg, 2013; Flanagin et al., 2006). Groups form as responses to issues on the political agenda or to promote specific political viewpoints. Examples of the expanding forms and repertoires of digital collective action range from hashtag-based global movements to small Facebook groups or one-shot events coordinated via social media. As organizations, such collectives are often considered more fluid forms of organizing that are somehow fostered by digital technologies. Scholars have investigated organizational forms and structures enabled by social media platforms (Dolata & Schrape, 2016; Ganesh & Stohl, 2021), the functions and uses of social media by movements (Myles et al., 2020; Theocharis et al., 2015; Tsatsou, 2018), and movement leadership supported by technology (Gerbaudo, 2012; Poell et al., 2016; Tsatsou, 2018). Recent scholarship also critically discusses the role of digital platforms as organizing agents and brokers (Bennett & Segerberg, 2013; Dolata & Schrape, 2016) whose properties and affordances affect the ways of performing digital activism (Kavada, 2015; Milan, 2015).

Existing research has often focused on named, well-known entities, such as Occupy (Theocharis et al., 2015), Arab Spring activists (Poell et al., 2016), or Anonymous (Dobusch & Schoeneborn, 2015), and explored the organizing of these collectives. This study adopts a smaller scale to study issue-based, seemingly spontaneous collectives that are commonly not regarded as organizations of any kind: informal discussion groups that convene online. In their classification of forms of collective action in the Internet age, Dolata and Schrape (2016) denote such constellations as non-organized collectives that aggregate individual action instead of acting strategically under a shared identity. As traditional face-to-face characteristics of social movements such as meetings (Haug, 2013) or street activism (Gerbaudo, 2012) are missing, the potential coordination of the collective must rely on technologically mediated communication, facilitated by the technological infrastructure of the social media platforms (Bennett & Segerberg, 2013). This makes organizing the collective into essentially a socio-technical process (Dolata & Schrape, 2016).

This study investigates the communicative, socio-technical organizing of an issue-based online collective, a large Finnish Facebook group related to urban planning and politics (hereinafter UPG).¹ UPG is profoundly digital: a public Facebook group with no formal membership. From the outside, it appears to be a grassroots-level collective for informal, non-organized discussions. However, the group has a stated purpose, it is known for well-informed and polite conversations, and it has received extensive external recognition by local politicians and the media (Niitamo & Sjöblom, 2018). With over 19,000 members, UPG is among the largest political Facebook groups in Finland. Can such a collective be self-maintained, or is it organized in some sense despite the absence of formal organization (Ahrne & Brunsson, 2011)? Guided by this thought, we conceptualize the group as an emerging digital social movement and analyze its organization-like nature. Our analysis focuses on an internal chat log of the group admins consisting of thousands of messages, which gives us a unique view of the movement backstage (Haug, 2013) and allows us to explore how exclusive admin communication unveils characteristics of a social movement organization.

By adopting a communication-centered ontological view to social movement organizations, our interest is in the organizationality (Dobusch & Schoeneborn, 2015; Schoeneborn & Dobusch, 2019) of digital collective action, that is, the degree to which these collectives engage in decision-making or generate a sense of collective actorhood and identity. We thus follow a communication perspective put forth in collective action studies (Flanagin et al., 2006; Kavada, 2015), which suggests that various social phenomena not typically viewed as organizations can nevertheless be explored through their communicative organizationality (Schoeneborn et al., 2019). To explicitly focus on communication and decision-making as highlighted in organizationality, this study builds on the premises of Luhmann’s Theory of Social Systems (TSS) and TSS-inspired literature within the Communicative Constitution of Organizations (CCO) approach in organization studies (Cooren et al., 2011). TSS proposes that organizations, as systems, emerge as they distinguish themselves from their environments through decision communication (Luhmann, 2006). Luhmann also asserts that organizations typically involve interactional communication between the people present in the interaction, which can be framed and conditioned by the organization (Seidl, 2005b). We propose that such an interwoven focus on decision communication and interactional communication works well as an analytical framework to explore the organizationality of a freely communicating online collective, where more institutionalized and material elements of organizing are often absent, but whose ultimate aim is nevertheless to be constituted and recognized as an actor representing a certain cause. This approach also helps to assess the potentially strategic and political nature of these groups. From these starting points, we ask:

RQ1: What characteristics of organizationality are present in the seemingly non-organized Facebook collective of UPG?

RQ2: In what ways do different forms of communication constitute the organizationality of an online movement and support its digital activism?

We now explain our conceptual premises for studying social movements as self-referential communicative constitutions and discuss their organizationality, then present our data and the qualitative analysis process. We present our findings by first discussing how UPG produces itself as an organization by reproducing organizational elements in decision communication and, second, how it aims to condition and orchestrate interactional communication within the group. Our results indicate that online groups are potentially more than informal sites of political expression; when strategically moderated, they can be covertly organized political movements. From the perspective of communicative organizationality, a combination of steering discussions by moderation and subtle nudging is a powerful practice of organizing people and discussions – a practice we call discursive conditioning.

Organizationality of social movements

Social movements are a form of collective action, informal networks based on shared beliefs and solidarity, seeking to promote or oppose social change and alter power relations (Della Porta & Diani, 2006). Movements often aim to transform societies through their engagement with the media, which makes communication a prerequisite of networking, sharing, and mobilization (Bennett et al., 2014; Loader, 2008). Recently, research has highlighted the role of digital media technologies in movement formation, suggesting that aided by technology, digital movements can adopt precarious and fluid forms, where memberships are vague and organizing occurs with weaker institutional cores and less coordination (Bennett & Segerberg, 2013; Dolata & Schrape, 2016; Kavada, 2015). Dolata and Schrape (2016) have classified different structures of online collectives ranging from digital crowds to communities and movements, drawing a distinction between collective actors with shared objectives, structures for strategic decision-making, and a separate collective identity versus non-organized collectives working to aggregate individual, spontaneous action without a core organization, postulated strategy, or collective decision-making. Issue-based digital movements, such as online discussion groups, are examples of the latter. They are perceived as loose, networked structures: emergent social orders that form autonomously around a common issue. Such collectives might, however, consolidate into more formal social movements over time (Dolata & Schrape, 2016).

More broadly, scholars have explored organizations that are not traditional or formal, but instead partial (Ahrne & Brunsson, 2011), fluid (Dobusch & Schoeneborn, 2015; Ganesh & Stohl, 2021; Schreyögg & Sydow, 2010), or proto organizations (Wilhoit & Kisselburgh, 2015). Studies have discussed how the new, emergent forms of organizing fulfill the criteria of organizationality, that is, the entity’s degree of organizational character (Dobusch & Schoeneborn, 2015) as well as the characteristics they lack. Ganesh and Stohl (2021) argue that the digital organizing environment especially enables and produces hybrid organizational forms. They discuss archetypes of digital organizing, from organizationally driven to organizationally embryonic types – with the main outcome of the organizationally embryonic archetype being digital visibility with nascent organizing structures and practices.

Dobusch and Schoeneborn (2015) define three criteria for organizationality: the degree to which the social collective displays interconnected instances of decision-making, is attributed actorhood, and makes identity claims regarding what it is or does. Others have also highlighted the role of decisions. For example, Ahrne and Brunsson (2011, p. 85) define organizations as ‘a decided order in which people use elements that are constitutive of formal organizations.’ These elements include membership (who is allowed to join), hierarchy (who makes decisions affecting others), rules for organization members, monitoring of these rules, and sanctions (allocating resources to members). Some organizational forms, however, can use only some of the elements of a formal organization (Ahrne & Brunsson, 2011). Scholars also argue for a gentle balance between form and formlessness: how to maintain organizationality without being bureaucratic, remain open and fluid without compromising boundaries and blending into the environment (Schoeneborn et al., 2019: Schreyögg & Sydow, 2010). It could be argued that to maintain such a balance, it is crucial to support flexibility in organizational practices and communicate in ways that support the management of coherence (Ganesh & Stohl, 2021). This calls for communication that expands the traditional repertoires of organizational communication in movements and that carefully harnesses the affordances of digital communication platforms – such as visuality or interactivity (Myles et al., 2020; Poell et al., 2016).

Social movements as communicative constitutions

For a social movement, a central way to gain power is to attract members or supporters and have those members act in coordination (Bennett et al., 2014). To accomplish this, communication is essential (Flanagin et al., 2006; Kavada, 2015). For a non-organized collective to consolidate and institutionalize its activities, resulting in more formal structures and goals, the emerging movement needs communication to create, modify, and sustain an organizational entity around the existing collective (Bennett & Segerberg, 2013; Kavada, 2015; Poell et al., 2016). Within organization studies, such thinking corresponds with the premises of the communicative constitution of organizations (CCO) approach, which considers organizations to be enacted through language use and its various manifestations, from documents to conversations (Cooren et al., 2011). Communication is, thus, not produced by or in an existing organization such as a movement; rather, it is the organization. Therefore, our interest shifts from the communication produced by a given entity to the emergence of the entity as an organization, originating from the micro-level communication in the collective (see Christensen & Cornelissen, 2011).

While not all communication in a collective is organizational, decisions are an essential component of organizationality. A CCO stream inspired by TSS emphasizes that organizational communication is communication connected to decisions and explores the constitution of an organization as a system of decision communication (Schoeneborn, 2011). In this view, organizations are social systems of autopoietic reproduction consisting of communications, and the network of communicative events – not the individuals or material objects – produces and reproduces decisions, actions, and the organization (Luhmann, 2006, 2018). Thus, it is not individual communicative events that matter, but a flow of communication, where the previous act of communication is interpreted in the next. The network of communication events then defines the meaning of each event, and the organization is reproduced through a chain of decision communication (Schoeneborn, 2011). For Luhmann, a decision is not only a choice among alternatives but a specific form of communication (Luhmann, 1992; Seidl, 2005a); decisions are communication, reproduced by the organization itself. At the same time, not all organizational communication is decision communication. Other forms of communications also exist in organizations and in their environment, such as gossip or purely interactional communication (Luhmann, 2018; Seidl, 2005a). These forms, however, do not contribute to organizational autopoiesis, only the chained decision communication does. Further, decision communication is contextual. Different decision premises in organizations define what decisions are produced: programs, personnel, and communication channels (Luhmann, 2018; Seidl, 2005a). Programs define the conditions for correct decision-making, personnel recruitment defines who is in charge of decisions, and communication channels premise who can communicate where and when.

Finally, for Luhmann (1986, 2005, 2018) the differentiation between the organization and its environment is an essential feature of the existence of the organization as a system. The differentiation takes place by maintaining boundaries, which means organizations are essentially boundary-maintaining systems (Hernes & Bakken, 2003; Schreyögg & Sydow, 2010). Luhmann emphasizes the communicative reproduction of the boundaries, a perspective also highlighted by CCO literature. Societal systems are operationally closed, but they communicate across subsystem boundaries as they are irritated by other systems; as the organization decides what to observe and what to communicate, it also reproduces itself by constituting a distinction between itself and the environment (Luhmann, 2018; Seidl, 2005a). Due to decision communication, organizations can communicate on behalf of particular concerns and be attributed as self-referential social actors (Dobusch & Schoeneborn, 2015).

Organizational and interactional communication

As mentioned, interactional communication, that is, communication amongst people, also takes place within the environment of the organization. While Luhmann asserts that organization and interaction are essentially separate systems, Seidl (2005b) notes that interactional communication can contribute to the reproduction of the organization. Interactional communications follow their own logic but also orient themselves according to organizational structures, which, in Luhmannian terms, is their environment (Seidl, 2005b). According to Luhmann, the organizational system affects the interactional system in three dimensions: social, factual, and temporal. Interaction reflects the social positions of the participants in the organization, organizational themes are reflected in the interactional themes, and finally, interactions often reflect the temporal structure of the organization, for instance by coinciding with organizational events such as meetings. Hence, the organization offers a frame for the interaction. Likewise, organizational interactions can act as decision premises for organizational decision communication. They can guide organizational decisions by offering more complexity and information, or they can be used as deparadoxifying elements if referred to by the decision-maker (Seidl, 2005b). Thus, systems of organization and interaction make their own communication and complexity available to the other system as elements for constructing the other (Luhmann, 2018; Seidl, 2005b).

As Seidl (2005b) maintains, organizations can condition organizational interaction in various ways beneficial to the organization. The organization’s system, however, has no direct control over the interaction because the interactional system itself selects its own premises and structures – the organization can only decide to stimulate or condition interactions. This, again, can occur in three dimensions: in the temporal dimension, the organization can schedule situations for participation and interactions; in the social dimension, the organization can decide who can participate and offer viewpoints to the interactions; and in the factual dimension, the organization can influence the selection of topics in the interaction (e.g., by setting the meeting agenda or by withholding information).

Building on these conceptualizations, we now empirically explore the forms of decision communication and conditioning in an informal online group.

Materials and methods

Our empirical case UPG is a Finnish Facebook group related to urban planning and city politics. UPG was founded in 2009 based on the personal political interests of its founder, along with a few like-minded acquaintances. According to the founder, the group started small. The number of members increased by roughly a factor of ten in 2013–2014, growing from a few hundred to 2,900 members by March 2014, possibly driven by media coverage. In November 2020, UPG had over 19,000 members. The group defines itself as a movement and discloses its goal on an abstract level. Its political relevance has been publicly acknowledged in the media and by city planning officers (Niitamo & Sjöblom, 2018). UPG has also hosted several live discussion events.

We conceptualize UPG as an emerging digital social movement, an example of issue-based citizen activism performed using the affordances of Facebook groups. Our data consists of the private Facebook chat logs of UPG admins between February 2013 and July 2014, a period characterized by the active growth of the group. The logs, given to the authors confidentially by the admins, include over 12,000 messages on 800 pages. They focus on the moderating actions and policies as well as general discussions about the activities, purpose, and future of UPG. They offer a window into the communication and negotiations taking place among the admins. Discussions were initiated almost daily, for example if an admin spotted relevant news or envisioned certain activities, and every time something in the public group required attention. Therefore, the conversations frequently referred to the public group discussion – which we conceptualize as organizational interaction.

The analysis was aided by Atlas.TI software. To obtain an overview of the data, in the first phase of the analysis the logs were carefully read and coded by the first author using a grounded, data-driven approach, modifying the codebook while going through the data by combining and re-naming codes when necessary. The aim of this round was to simply understand what is going on in the data (Gioia et al., 2013), without being limited to theoretical perspectives. However, to remain faithful to our research problem, the first round of coding focused on passages in which the admins discussed organization-related issues, such as UPG activities, purposes, future goals, and communication practices. Topical discussions about city politics or news or general chatter were not coded. After this phase, we ended up with 31 grounded categories (Table 1). Codes were allowed to overlap. During this initial exploration, we were surprised to observe how much the admins strategized about their communicative interventions in the public group and also how frequently and in what ways they talked about the regular group members.

Table 1. Grounded codes organized according to the theoretical framework.

Theoretical dimension			Grounded codes
Organizationality (Dobusch & Schoeneborn, 2015)	Elements of decision making (Ahrne & Brunsson, 2011)	Negotiating membership	Tracking joining members Restricting membership Lack of women
		Hierarchy building	Educating members Non-politicians Personal political goals Notable persons Differentiating from ordinary members Criticizing members Mobilization Inclusion
		Sanctions	Formal warnings Banning Rewarding members
		Rules	Reference to rules Moderation
		Monitoring	Moderation Promoting online conversation skills Disqualification of oneself
	Actorhood		Media coverage Group reputation Group legitimacy Hype and back-patting Organizing events Concrete doing External strategic action External communication
	Identity / Boundaries		Reference to rules Organizational identity Group mission Hype and back-patting Visionary talk
Interaction	Discursive Conditioning (Seidl, 2005b)		Conversation culture argument Internal strategic action Moderation Promoting online conversation skills

In the second stage, both authors used the coding to organize the data from the perspective of organizationality (RQ1). The literature was consulted, after which the grounded codes were juxtaposed with the organizational elements (membership, hierarchies, rules, monitoring the rules, sanctions) (Ahrne & Brunsson, 2011) as well as the dimensions of organizationality (decisions, actorhood, identity) (Dobusch & Schoeneborn, 2015) in a hermeneutic process to structure the data (Gioia et al., 2013). Some grounded codes corresponded with multiple theoretical dimensions: Deleting comments was considered an example of monitoring the rules and giving warnings to members an example of imposing sanctions. Then again, the concurrent discourse on hype and back-patting served as both actorhood and identity claims since they were often triggered by attributions of actorhood, followed by statements defining the excellence of the group. Thus, the structure presented in Table 1 is not a formative, deductive data structure, but rather an attempt to abductively organize the data using the concept of organizationality, while also revealing potential inconsistencies and anomalies (Locke et al., 2008).

As a result of the structuring, we indeed encountered a peculiar anomaly: while our exploration had started from the perspective of communicative organizationality, our original coding included prominent codes for communicative operations that could not be categorized with our theoretical framework. Therefore, we grouped together a set of four grounded codes (conversation culture argument, internal strategic action, moderation practices, promoting online conversation skills). These practices highlighted the strategic nature of how admins performed communication within the public group, a finding supporting existing research on protest organizing and leadership as well as Luhmann’s ideas of conditioning organizational interactions. This led us to refine our second research question and shift the focus of our exploration to the different forms of communication and their role in constituting the online movement. In the final stage of the analysis, we revisited the data to describe how these communication-related practices help constitute the organizationality of the emergent movement (RQ2). The first author re-read the passages marked with the communication-related codes and contrasted them with the forms of conditioning organizational interaction (Seidl, 2005b). While these codes described practices connected with the social, factual, and temporal dimensions of interaction, they also emphasized a subtler way to intervene in the group discussions in order to sway discussion topics and promote a certain culture of online conversation. Inspired by the TSS framework, we called this dimension discursive conditioning of interaction and assessed its practices, including the elements of organizationality, using the interpenetration of organizational and interactional systems. In the results section, we discuss the activities and communications of the group using the categorization shown in Table 1 as our reporting guide, while reflecting on the findings from the perspective of conditioning interactions throughout the analysis.

Tracing the organizationality

The public Facebook group displays political but informal discussions – an example of nascent organizational structures (Ganesh & Stohl, 2021). Our empirical materials originates in the backstage of these discussions, where a group of admins discussed, for instance, UPG’s vision, function, members, and actions. Their chat interchangeably hosted both an interaction system and decision-making organizational system. The excerpt below shows a typical communication episode in the logs: an admin points out a matter for consideration for the public group, and the participants then reach a decision:

B: we have [name] now presenting some basic citizen racism

B: [pastes the original comment]

B: Slightly out of limits. Fine as a single comment, but the way he hints at unjustified generalizations doesn’t fit with our conversation culture –

C: For some reason, I cannot touch that post. I don’t know if [person] blocked me or something, but that could be removed. It’s big bad R[acism] itself.

D: Send him a private message and argue that since this is a forum for city politics, we don’t want the shouting and fighting that follows.

(October 2013)

Such conversations are frequently initiated because something in the group catches the admins’ attention or in response to external events that are relevant for UPG’s political purpose. Thus, the admin chat logs reveal organizational self-observation and subsequent interconnected instances of decision-making, which is deemed one of the main characteristics of organizationality and organizational systems (Dobusch & Schoeneborn, 2015; Luhmann, 2018). Ultimately, each new decision contributed to an unwritten understanding about conversation topics and styles allowed in the group. The results of the decision communication were communicated to the public group sometimes openly through moderation, or as speech acts masked as regular interactions. In this way, the outcomes of the decision communication were made available by the organizational system as triggers to the interactional system inside and overlapping with the larger system of the Facebook group (Seidl, 2005b).

In what follows, we explore decision communication concerning organizational elements in UPG, then discuss the empirical practices related to actorhood and group identity, and finally show how the organizational system sought to condition interactions in the public group.

Decisions on organizational elements

First, decisions targeted membership negotiations and hierarchies. Technically, membership negotiations for a public Facebook group seem minimal: joining members are accepted by admins, but this procedure is mostly a precaution towards fake users. Explicit and implicit hierarchies do, however, exist among members. Explicitly, members are divided into administrators and regular members, which is also structured by the platform technology. The names of admins are publicly declared, but no formal leadership is declared. The admin role, nonetheless, extends beyond technical administration and spam moderation: they formed an ideological core for the group and formulated its original mission. Thus, some forms of non-hierarchical leadership shown to exist in digital movements were present (Gerbaudo, 2012; Poell et al., 2016; Tsatsou, 2018).

Further, tacit hierarchies extended also beyond admins to regular members through their discussions, evaluating members based on their political or professional position or communicative skills and activities (Ashuri & Bar-Ilan, 2016); members’ actions were constantly weighed against the overall purpose of the group. These discussions reflected a desire to expand the movement’s visibility beyond the online conversations, to build a collective we (Milan, 2015). Notable people such as politicians, municipal officials, and journalists were especially appreciated, as were people with specific knowledge or skills, for example in research:

B: [Person] (who just took part in the [thread] and seems to know about city economics) is a researcher at [university], studies urban development and transport systems – seems to be a valuable addition to the group. (October 2013)

Thus, members were valued using instrumental evaluations in relation to UPG’s goal. While invisible, these evaluations acted as concrete decision premises that affected the decisions of the admins, for example who were invited to contribute to organizing events or writing statements. These are communicative events that help constitute member groups and membership hierarchies. In Luhmannian terms, they are forms of social conditioning of organizational interaction; the admins tried to influence who could participate and what perspectives to include.

Another decision-making premise for conditioning the interaction is the general policy of conversation rules published for UPG. The policy includes the expressly stated purpose of the group, general conversation rules, and moderation policy. Apart from being an abstract representation of UPG and its collective intentions (Cooren et al., 2011), the document was used as a code of conduct to legitimize admins’ actions. Yet, the majority of admin discussions focused on monitoring the rules, discussing the severity of potential violations, and deciding on consequences – following the decision-making program (Luhmann, 2018) explained in the policy document. For instance, discussants were expected to stick to the topic, use references to justify their arguments, refrain from verbally attacking other members, and avoid irony. All these meta-conversational aspects are an important part of the online conversation skills that the admins consider a trademark of UPG.

Failure to follow the written and unwritten rules of the group will lead to sanctions. Light forms of sanctions are what the admins call ‘formal warnings,’ almost always accompanied by the removal of the related post or comment. The most extreme form of sanctioning is restricting membership – conditioning both the social and temporal dimensions of interaction. If a member has been given multiple warnings, they are banned temporarily or permanently. All forms of sanction are communicated to the target person using private messages. Furthermore, the admins work to develop positive sanctions. For example, in the summer of 2013 a special symbol for awarding members was invented: penguins awarded virtually or as fluffy toys. Awarding is a way of educating members: a successful process of socializing the discussants to UPG’s culture and to the ways of doing proper pro-urban activism, thereby earning them an award:

A: we should start awarding with reputation and honor those who have actively worked for the cause - - an incentive for newbies to start doing things and a reminder that this is not only a conversation club, but we actually do things here. (July 2013)

Constituting identities, boundaries, and collective actorhood

Research suggests that organizational identity must be continuously negotiated and that there exist competing conceptualizations of an organization’s identity (Dobusch & Schoeneborn, 2015). Such negotiations are further highlighted in contemporary digital forms of activism, where individual performances become central to collective identity (Bennett & Segerberg, 2013; Milan, 2015). While differing conceptualizations likely exist among UPG members, the admins seemingly share a unified understanding of their group’s identity and purpose – petrified in the rules document – and a hubris-like agreement on UPG’s political justification and excellence. They use the pronoun we frequently but ambiguously. Sometimes it refers to the public collective (‘we have so much buzz that we need space for action, active people should get to know each other’), sometimes it refers to the admins only (‘we need a process to audit people we don’t know that well’). The hybrid hubris identity is profoundly guided by admins’ pro-urban values and a firm belief in their cause:

C: My strategic viewpoint has been, all along, that our time has come, and our cause is good, right and noble, and all we have to do is shove the sleigh off, and things will start to happen. (March 2014)

There are only a few explicit discussions of identity, where the admins discuss their practices of organizational voicing – moments when the organization observes itself as a participant (Luhmann, 2018). However, the admins constantly express concern about UPG’s external reputation and legitimacy; they rigorously follow external statements about the group, accompanied by an urge to position UPG as an actor. King and colleagues (2010) suggest that organizations can be differentiated from other forms of collective behavior via external attribution: they are perceived as actors by others in society. Through the strategy discussions and planning of external communication, the admins aim to position UPG itself as an actor in the political field, but they crucially need their regular members to build public credibility and visibility (Milan, 2015). The strategy works; during our study period, referrals to external acknowledgment become more frequent:

B: [links a column that mentions UPG]

B. Let’s see if that causes another invasion to the group

B: Luckily, it seems potential newcomers will already be inclined towards the correct conversation culture.

(August 2013)

Awareness of the growing public traction gained by the group fosters an ironic discourse of hype and back-patting: ‘We command and govern. The whole city is ours’ (August 2013). In this way, UPG seeks to become a full-fledged movement; an organization is formed when it has a voice and perceived as a self-referential actor (Dobusch & Schoeneborn, 2015).

Conditioning interaction by nudging

Moderating UPG takes subtler forms than merely deleting unsuitable content and banning members. The admins nudge the discussion in the right direction through strategic discussion moves aligned with UPG’s goals: by opening new threads at strategic times on current issues in the news or on the city council’s agenda. Additionally, they post planned comments to existing threads either to fine-tune the discussion or to elevate the thread in the newsfeed. Sometimes, they post strategic comments to encourage responses from valued members. They essentially attempt to condition organizational interaction in all three dimensions (Seidl, 2005b), but in a less commanding way than, for instance, via meeting agendas. To some extent, the ‘right direction’ of discussions advocated by the admins is explicated in the rules, where the group’s purpose is spelled out, but also defined contextually, case by case, in chains of decisions communicated by the admins:

C: An excellent opening on arguing from a business perspective. Can we add a rule that general contradicting is forbidden, and you can do that elsewhere?

A: I wrote a short one.

A: Throw in some viewpoints to start with, if the conversation doesn’t otherwise go in the right direction?

(December 2013)

Further, as shown in previous quotations, when moderating the discussions, the admins constantly highlight the importance of the online conversation skills and conversation culture mentioned in the rules. Both concepts advocate a fact-based, rational discussion style. In fact, guarding the conversation culture is seemingly the main purpose of the rules and a second ideological premise of UPG. It forms a composite with the pro-urban ideology, and it is frequently mobilized to help regulate discussion in the public group. Simultaneously, the meaning of the conversation culture is constantly contested and constructed in the admin discussions, and the admins repeatedly discuss and adjust their actions beyond what is postulated in the rules:

A: C, if you knew that person, could you kindly suggest that (s)he wouldn’t open threads for a few days and consider the conversation culture a bit. (August 2013)

Thus, our data highlights the pronounced nature of the admins’ communication and underlines their hidden political goals. Through planned new threads, subtle comments, or strategic likes, the admins carefully marshal the topics and rhythm of communication in the public group. For regular members, the decision communication behind these delicate acts is visible only through moderation or orchestrated communication in the public group – communication that the regular members might regard as part of the interaction. We term this process insinuated, discursive conditioning, that is, influencing organizational interaction by nudging the conversations in the emergent movement in preferred directions.

Discussion

In this study, we explored how the organizationality of a political Facebook group is constituted in communication. By tracing the characteristics of organizationality, we demonstrated the existence of decision communication in the admin discussions and how their subsequent decisions regarding organizational elements are manifested in public group interactions. Building on Luhmann’s TSS, we analyzed how organizational interaction is conditioned (Seidl, 2005b) using organizational elements and subtle, strategic communication planned beneath the façade of the putatively non-organized group.

In the social dimension of conditioning, the admins control who to approve as members and sometimes revoke membership. They construct invisible hierarchies by judging the perceived usefulness of members, followed by decisions to encourage more valued members to become more active in the group through invitations, awarding practices, and even supportive public interactions. Through these means, the organizational system uses personnel choices to restrict viewpoints in interactional communication. In the context of Facebook, this has direct consequences: certain people are favored to produce a desired political discussion within the group. Thus, the social conditioning of interaction fosters factual conditioning. There is little direct control solely in the factual dimension. However, the admins sometimes use platform-afforded practices of content moderation and their admin status to steer the topics of communication in the interaction. The factual conditioning also includes opening discussion threads and posting comments to influence the prominence of threads in the group newsfeed.

Conditioning the temporal dimension in the digital context is a less-structured process than in a ‘regular’ organization, since in principle the group is open to interaction at all times. The admins, however, work to condition the interaction temporally in two ways: by adjusting the temporal rhythm of topics and by silencing and activating members. First, they strategically make certain topics available to the interaction in accordance with local or national political agenda – factual and temporal conditioning become intertwined. Second, the admins sometimes decide to temporarily silence members through bans, or they terminate interaction by closing heated threads.

Finally, when exploring the communicative practices of the admins, we identified a form of conditioning organizational interaction that is profoundly communicative, overlapping and intertwined with the other forms: discursive conditioning. Luhmann theorized about how organizations modify contextual factors (social, factual, temporal) to condition organizational interaction by stimulating the interaction system in various ways (Seidl, 2005b). Discursive conditioning does not deal with the context of the interaction but uses strategic, obfuscated speech acts within the interaction to nudge the conversations in desired directions – both in relation to discussion topics and conversation style. In UPG, the admins continuously mobilize the ambiguous notion of conversation culture to guide members’ interaction style. It is an ideological element proudly defended by the admins, and a powerful, yet subtle, way to condition the discussions by indirectly intervening with the factual dimension through style-related arguments. The use of it is thus politically expedient, even if perhaps not intended.

The unique empirical observation of this study is the insinuated combination of factual and discursive conditioning, manifested as communicative acts that aim to stimulate more interactional communication. The acts of discursive conditioning are performed in the interactional system, but they are outcomes of decision communication in the hidden organizational system: hybrid acts of communication that work at the boundaries of the two systems, yet still following the differing logics of each system. In a regular organization, interaction could be conditioned through meetings and meeting agendas, which are recognized as organizational. The discursive conditioning observed in our data is a subtler, even obscure way to stimulate the interaction. The admins’ speech acts in the interaction system are shared by the two systems, but they remain unmarked and cannot be identified by the other system (Seidl, 2005b).

Essentially, conditioning is made possible by the technological platform, which supports the penetration of the two systems in novel ways. Some ‘regular’ ways of conditioning interaction, such as limiting participants by bans, are directly afforded by Facebook’s programed features. From the perspective of temporality, the overlapping organizational arenas created by the platform allow the admins to strategize their conditioning in real time while observing the organizational interaction. The combination of factual and discursive conditioning is also a careful play with the platform features: the admins use the logic of the newsfeed to promote threads in the group. Thus, the organizational structures are intertwined with the technological structures of the platform, and conditioning finds new forms when fused with the techno-social practice of moderating groups.

It is notable that while Facebook offers straightforward ways to control discussion topics, the admins have nevertheless chosen to condition organizational interaction in subtle ways by nudging and promoting the conversation culture. Such conditioning resembles forms of strategic communication designed to support the organization’s mission, but not targeting stakeholders – rather, as suggested by Van Ruler (2018), as the strategic management of communication flows, where strategy is not only promoted and built, but also refined and negotiated through communication. Further, as highlighted by Christensen and Cornelissen (2011), the CCO/TSS lens shifted our focus from sender-centered transmission to constitution: discursive conditioning is a form of communication that constitutes the organization. Most importantly, we argue that covert, discursive conditioning is an essential mechanism for scaling an online group up to a movement organization while preserving the engagement of the members and creating a sense of shared identity. By delicately stimulating the public organizational interaction within the group, the organizational system produces desired outcomes through digital activism. For a digital social movement organized as an issue-based group, the public discussions constitute the main ‘product’ of the movement, which disseminate the political views of the group and potentially influence the political system. The conditioned interaction orchestrated by the admins appears as a form of governance used to advance their organization’s political agenda – a seemingly non-organized collective emerges as a substitute for its founders’ political statement.

The resulting movement is a hybrid form of online collective action (Bennett & Segerberg, 2013; Dolata & Schrape, 2016): externally, it retains the form of a non-organized, technology-enabled collective, but aided by different forms of conditioning, the admins build recognition and actorhood for the emerging movement organization, thus sanctioning a simultaneous yet disconnected organizational state for the movement. We suggest that this is a strategic choice; a low degree of visible organizationality is advantageous for them (Ahrne & Brunsson, 2011; Schoeneborn & Dobusch, 2019). As existing studies highlight, softer, connective forms of guiding activism seem to be well supported by online platforms (Gerbaudo, 2012; Poell et al., 2016). Coleman (2014) proposes that movements tend to present themselves as collectives, while hiding the leaders and their potentially problematic political affiliations. Yet, guided communication is essential to condition the interaction towards a common political vision. With UPG, all intertwined forms of conditioning work to build organizational actorhood for the group, with it being increasingly acknowledged for topical, high-quality discussions. Thus, the identified communicative forms of organizing are, somewhat contradictorily, used to organize discussions instead of people, even though people presumably form the core of connective action (e.g., Bennett & Segerberg, 2013; Poell et al., 2016).

Conclusion

This study applied Luhmannian organization theory to explore the organizationality of a seemingly non-organized Facebook group and showed how an issue-based online collective can be covertly organized by conditioning interaction. As such, the study makes three contributions. First, it stresses the complex nature of the constitution of online activism by showing that seemingly unorganized discussion groups might be organized even if they resemble networked, fluid, or embryonic action (Ganesh & Stohl, 2021). While studies have shown that organizational elements exist in online communities (e.g., Dolata & Schrape, 2016), our findings highlight their ambivalent visibility, afforded by the platform technology. Here, the distinction between organization and interaction systems (Luhmann, 2018) provides conceptual tools to theorize about and explore forms of organizationality in online movements, and in particular the role of strategic communication in them. It directs our attention to ways in which public online discussion – interaction – can be conditioned and stimulated by the – perhaps hidden – organizational system and how strategic, insinuated acts of communication steer further communication in interpenetrated systems, thereby constituting the organization.

Second, we contribute to Luhmann-inspired organization studies by applying the organization-interaction distinction in the context of digital organizing instead of traditional, bureaucratic organizations. Both interconnected forms of communication contribute to the constitution of the UPG group as an organization. Further, in this context we identified a novel practice of conditioning organizational interaction: subtle, discursive conditioning of topics and styles of interaction, a communicative practice that highlights the potential power of strategic communication in steering public conversations. Through hybrid and unmarked speech acts, the organizational system makes itself covertly available to the interaction system to push the interaction in directions beneficial for the organization. Our work also foregrounds the intertwined nature of the different dimensions of conditioning interaction. In the case of digital activism, where the main objective is to produce desired political talk, discursive conditioning is vital, but simultaneously intertwined with, supports and is supported by previously identified forms of conditioning. Finally, this observation also connects theories of strategic communication and organizational communication by showing how purposeful, strategic acts of communication also constitute the organization: the organization is not the sender of strategic communication, but rather the outcome (see Christensen & Cornelissen, 2011). This implies that there are forms of strategic communication not commonly identified as such by the scholarship on strategic communication (also Van Ruler, 2018).

Third, our findings contribute to online moderation studies by showing how moderation practices and the affordances of digital platforms can be used to produce and curate online political discussions and activism without heavy promotional activities, staged identity, or communication mechanisms like personal action frames (Bennett & Segerberg, 2013; Milan, 2015). Instead, digital activism can employ subtle, technology-assisted ways of thematically and discursively conditioning online interaction to shape public political talk. Discussions on moderation have focused on removing harmful content from platforms (e.g., Gillespie, 2018). Our findings pinpoint that moderation can pursue more diverse purposes and goals (also Ruckenstein & Turunen, 2019). By applying a Luhmannian framework, we conceptualized moderation practices as a form of conditioning organizational interaction (Seidl, 2005b). While Facebook is a temporally and spatially open environment for interaction, its technological affordances built for moderating groups also support forms of conditioning political online discussions when combined with strategically nudging conversations.

To conclude, this study highlights the importance of studying communication as the constitutive force of social movement organizations, particularly in the socio-technical context of digital platforms. The idea of discursive conditioning highlights how orchestrating public online discussions can be a powerful tool to propel networked digital activism or foster digital political expression. However, some limitations are generated by the admin-focused data. While traces of the public discussion are present in the chats, an analysis of the public group content would give a more comprehensive picture of the organizing process. Further, our findings are dependent on the national, political, and cultural contexts of using digital platforms. Similar political influence by a Facebook group might not be achieved in a country with lower Internet penetration and decreased general trust in online discussions. Group moderation practices and affordances, however, are institutionalizing globally, and it is reasonable to assume that other groups or actors are similarly using conditioning strategies to achieve their political goals. We encourage future studies to examine the work of unprofessional group moderators and the potential political and social implications of their practices.

Acknowledgements

The authors would like to thank the anonymous informants for generously giving access to their data and insights, and the two anonymous reviewers for their detailed and dedicated comments that significantly improved the manuscript. We are also grateful for Anu Harju, Ella Lillqvist as well as other colleagues in EGOS and ICA conferences for their comments on previous versions of this work .

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Salla-Maaria Laaksonen

Salla-Maaria Laaksonen (D.Soc.Sc.) is a postdoctoral researcher in the Centre for Consumer Society Research at the University of Helsinki. Her research areas are technology, organizations and new media; including narrating organizational reputation in the hybrid media system, the organization of social movements in the online context, and the use of data and algorithms in organizations. Further, she is an expert of digital and computational methods.

Merja Porttikivi

Merja Porttikivi (D.Econ.Sc.) is currently a researcher at the Department of Management Studies at the Aalto University Business School. Her research areas fall at the intersection of communication and organization studies; including organizational legitimation in the online public sphere and discursive approaches to social media.

Notes

1 Group name is pseudonymized to secure the anonymity of our informants.