One of our goals at Open Future is to make sense of openness and the movement advancing this principle. To this end, we have conducted an exploratory mapping of the movement using network analysis methods and data collected from Twitter.
In the past two decades, the open movement has been continuously working towards a digital future that is more democratic — it has been a driving force behind initiatives that have contributed to the democratization of knowledge and information.
Therefore, it is critical to continuously assess the role and relevance of openness in the context of evolving technological landscapes and identify new policy opportunities and challenges for promoting openness in ways that prioritize public interest values and benefit society.
One of our goals at Open Future is to make sense of openness and the movement advancing this principle. We aim to understand how the achievements of the last two decades can be stewarded and maintained and the new, emergent issues and challenges. We aim to understand shifting strategies and advocacy goals. Finally, we are exploring issues around openness and power imbalances, often raised as a fundamental challenge to openness, as it has been framed in the last two decades.
This work requires understanding openness and its advocates' shifting goals and strategies. It also requires defining and mapping the open movement. Who is part of this movement, and how is it structured?
To this end, we have conducted two studies in parallel. One of them — the one you are reading — is an exploratory mapping of the movement using network analysis methods and data collected from Twitter. The other is a qualitative survey of open movement leaders, titled "Shifting Tides: the Open Movement at a Turning Point."
This research brief presents the results of the mapping study. The first part consists of a definition of the open movement, followed by a conceptualization of the movement as consisting of distinct but connected “fields of open”. The second part comprises methodological information about social network analysis and data visualization. The third part includes a presentation of the network visualizations and their analysis.
As our network analysis confirms, one of our main findings was that the movement is divided into interconnected fields of open activism that have developed over at least the last two decades (with some of the fields having an even more extended history). The analysis provides insights concerning movement cohesion and proves helpful for developing movement-wide strategies and advocacy goals.
The data visualization work for this study was conducted by panGenerator. Data collection and analysis was done by Alek Tarkowski, Francesco Vogelezang and Kasia Fantoni.
Over the last two years, Open Future has been exploring the future of openness: talking with activists, experts, and leaders of organizations in the open movement about the future of openness. Our research unearthed a heterogeneity of viewpoints, often stemming from differences in context.
There is also a continuing ideological debate (albeit not as heated as in the past) about the term “open,” its meaning, and possible alternative terms: Access to Knowledge, free culture, free knowledge, Open Access, peer-to-peer, sharing, and the commons. Different activists and organizations try to make sense of the current state of openness — we have explored these challenges through a conversation about the Paradox of Open.
As Open Future aims to work for the open movement, we have been asking ourselves: What exactly is this movement? How can it be defined? And what are its boundaries?
Open activism, in many ways, fits the concept of a social movement. Renowned sociologist Arjun Appadurai describes a social movement as a heterogeneous group of actors tied together by a common identity, shared meanings, or collective imaginarium.1 The definition stresses that movements deal not only with material stakes or economic growth but also with values and cultural change.
French sociologist Alain Touraine suggested that social movements aim to change significant cultural orientations of our societies.2 One could definitely have that sense when participating in open activism in the first decade of the century, as it aimed to shape key models of production and intellectual property regimes in our knowledge-based societies.
These definitions help capture the specificity of collective action in support of openness, structured neither through market relations nor state intervention. Instead, the activity has a decentralized, networked character.
At the same time, our sensemaking study with open movement leaders shows uncertainty about whether openness is indeed championed today by a social movement. There is an overall sense of a shared identity, values, and ties that link people and organizations into networks. Yet the strength of these connections is not necessarily sufficient to constitute a social movement.
Nevertheless, for the purpose of this mapping study, we will use the term “open movement” to describe the overall structure we aimed to map and understand.
While we recognize that there are multiple, varied, and sometimes contesting understandings of openness and open activism, we needed a clear definition of the open movement for this mapping study.
We decided to define the open movement in the following way:
The open movement consists of people, communities, and organizations who (1) contribute to shared resources online that are available for everyone to use and reuse, (2) and/or advocate for non-exclusive access and use of information resources.
We do not aim to solve the issue of different ideologies and contested terminologies by providing this definition. Instead, we use it to help us identify who belongs to the movement and what its boundaries are.
This definition aims to encompass the two main strands of open activism. One is defined by engagement in productive work, which can broadly be described as fitting Yochai Benkler’s concepts of social production and commons-based peer production.3 The other relates to Touraine’s idea of a social movement as engaged in political activities: in this case, the politics of open.
The definition also seeks to define openness in a way that’s not recursive and goes beyond stating that “open activists promote openness.” Instead, it frames openness regarding “non-exclusive access and use of information resources.” This part of the definition is based on the language proposed by the Open Definition, stewarded by the Open Knowledge Foundation:
“Knowledge is open if anyone is free to access, use, modify, and share it — subject, at most, to measures that preserve provenance and openness.”
In a recent piece written for our Paradox of Open series of essays, James Boyle provides a slightly different conceptualization of the open movement as consisting of two intersecting circles. The first is “the movement to enable a series of privately created sharing commons.” Boyle points out that this movement ranges from activists aiming to reshape the economy to those for whom “open licenses were merely tools to solve some particular task” – mirroring the distinction between advocacy and production of resources in our definition.
According to Boyle, the second part of the open movement was “built around the fight for the open web”: a decentralized open global communications architecture with layers functioning as common property. According to Boyle, the second type of open activism was galvanized when a threat to the open web came from copyright policy, more precisely in the debates on intermediary liability. There was also an earlier, productive form of open web activism when the open web was designed and built. Our definition highlights this dual character of open web activism.
We can therefore conceptualize the open movement as an ecosystem that can be mapped on two axes. One is defined by the dominant activity type: advocacy on openness or production and maintenance. The other axis concerns the type of resource to which this activity applies: different types of information resources or the web itself.
The open movement can be further conceptualized, in more detail, by thinking in terms of different but interconnected fields of open.
Each of these fields consists of a domain-specific application of the general principle of openness: that non-exclusive access and use of information resources is better than claiming exclusive ownership over these resources regarding social and economic outcomes.
Fields of open share a common denominator: some variation of open licensing (such as the Creative Commons licenses), or reliance on the public domain status of resources, is a cross-cutting feature of these fields.
By distinguishing these open fields, we acknowledge the historical development of the open movement, which occurred through applying principles and tools for openness to different spheres of society and different resources. Free software programming and Open Access in science were among the earliest fields to emerge, followed by such new fields as Open Education, Open Data, or Open GLAM (also called Open Culture).
Admittedly, the distinctions between fields are not clear-cut. One example is the distinction between a more specific yet firmly established field of Open Access (focused on scientific journals) and the broader field of Open Science. Another example is Open Data, which overlaps with two other fields: Open Science and Open Government.
Some fields have also been defined by foundational documents or manifestos adopted by stakeholders in the given field: the Budapest Open Access Initiative, the Cape Town Open Education Declaration, Open GLAM Principles, or the Open Data Charter. Others, like the field of free culture (defined largely by Lawrence Lessig’s eponymous book), have less clear domain boundaries. They can nevertheless be identified by a common purpose or the societal challenge they are addressing.
Thinking in terms of fields is helpful in understanding the differences between them and, thus, differences in approaches to openness. Each of these fields could be understood as a social movement of its own, focused on a particular sphere of life and associated set of resources.
Each field has a distinct “open stack”: a set of tools, mechanisms, methods, protocols, norms, and infrastructures that define how the general open norm is expressed in a given field. For example, Open Education activists have been paying a lot of attention to the agency of users (educators and learners) in a way that is uncommon in other fields.
There are also differences in licensing standards; even if the same, generic licenses are a tool used across field boundaries. To continue the example of Open Education, non-commercial licensing clauses have gained popularity in this field to an extent not seen in others. Naturally, licensing standards also differ due to characteristics of resources — for example, the field of Open Data is strongly influenced by the fact that data is not copyrightable.
These domain-bound fields are not the only aggregates within the open movement. For example, large movement organizations build around themselves ecosystems that are in many ways similar to the fields of open. Wikimedia, Mozilla, and Creative Commons (at a lesser scale) have all, to some extent, created networks and ecosystems that cut across boundaries of different fields. And that aims to engage activists around the organization’s mission, sometimes even brand. Yet even these organizations usually define their work with reference to one of the fields of open. For example, Mozilla has always been working on Open Web advocacy and currently has a strong program connected to Open AI. Creative Commons focuses on Open Culture, Open Education, Open Climate, and Open AI.
Through our analysis, the following fields of open can be identified:
Free Culture: Organizations and individuals working on opening cultural and creative production. Free Culture activism encourages original creators to share their works under conditions that allow reuse;
Free / Open Source software: Organizations and individuals promoting the use and creation of Free and Open Source Software through free and/or open licensing of broadly understood code;
Open Access: Organizations and individuals working on opening up academic publications (or more general academic outputs) based on the principles defined in the Budapest Open Access Initiative;
Open Data: Organizations and individuals working on opening up data; this is a field that cuts across domains, with a solid connection to both Open Science and Open Government;
Open Design: Organizations and individuals promoting the use and creation of physical products through the use of publicly shared and openly licensed design information;
Open Education: Organizations and individuals working on opening educational resources, as well as other elements or aspects of education systems and processes, based on the principles defined in the Cape Town Open Education Declaration;
Open GLAM / Open Culture: Organizations and individuals working on opening the collections of cultural heritage institutions. GLAM stands for Galleries, Archives, Libraries, and Museums. Also sometimes called Open Heritage or Open Culture;
Open Government: Organizations and individuals working on the opening up of data and information held by governments, usually with the aim of improving government transparency and civic participation;
Open Hardware: Organizations and individuals promoting the use and creation of Open Source hardware, through the use of publicly shared and openly licensed technical specifications;
Open Internet / Open Web: Organizations and individuals working to secure the open character of the web / the internet, at the different layers of the internet stack;
Open Science: Organizations and individuals working more broadly on opening up scientific research.
The field of Open Internet or Open Web activism is an outlier in this regard. In this field, openness is applied not to information resources but to the architecture of a communications infrastructure: the internet. At the same time, this infrastructure (with other innovations, such as new licensing models or new societal norms and modes of social production) enabled other fields to emerge.
For the purpose of our research, we also identified an additional category of movement actors that do not fit neatly into the model of the fields of open: organizations and people working on copyright reform advocacy. Our network analysis shows that this group is close to the field of Free Culture in terms of its position in the movement — although copyright reform is not limited to this field. Ultimately, we consider copyright reform activism as transversal to the fields of open.
Finally, it’s worth noting that the concept of the fields of open captures not only well-established, incumbent parts of the movement but also its emergent parts. These are often still in flux regarding key actors, strategy, or the specific version of the stack of open tools being deployed. We nevertheless see value in identifying such emergent fields. One current example is Open Climate activism, which aims to open up environmental and climate data and research used to combat the climate crisis. Another is the emerging field of Open AI: constituted by organizations and individuals opening up machine learning systems. While the status of such fields might not yet be clear, they have the value of offering new perspectives on openness, and new means of achieving it, to the broader movement.
To obtain a more comprehensive understanding of the open movement as an ecosystem, we decided to use Twitter as a source of information on who belongs to the open movement and their patterns of interaction. We refer to this part of the open movement inquiry as "open movement mapping." Despite the name, however, we are not claiming that the results of our research present a complete map of the open movement. Our ambition is much more modest – we want to glimpse how the ecosystem was reflected in the Twitter-sphere at a given time.
When we collected the data, we were aware of Twitter's volatile nature and the uncertainty surrounding its future. Towards the end of 2022 and the beginning of 2023, a pervasive sentiment emerged that Twitter, in its current form, may be approaching its end. We realized that the data we were analyzing might be Twitter's users’ swan songs.
On the other hand, we still were confident that the interactions on Twitter between members of the open movement are a useful, albeit partial, source of insight into the structure of the open movement. Finally, data made available by Twitter through its API is unique because it offers an opportunity to analyze and map the open movement quantitatively. We recognized the value of studying Twitter data as it allowed us to capture a snapshot of the ecosystem at this interesting juncture in Twitter's history.
Attempts to use social network analysis to study the structure of a social movement have several limitations that must be considered. Significant limitations to our inquiry included:
Bias in the data and the risk of contextual inaccuracy — Twitter only provides information on what occurs on Twitter. Twitter data does not represent the entire movement or even capture the full range of communication activities occurring on other networks. Its analysis does not capture the full range of online activities and interactions of social movement members. As a result, analyzing Twitter data in isolation may lead to flawed conclusions.
Noise in the data — Twitter data may contain spam, bots, and irrelevant information, making identifying and analyzing a social movement's network structure difficult.
Data access restrictions — Twitter's API restricts data access, limiting the depth and breadth of data researchers can analyze.
And in general, network structure analysis can lead to deterministic explanations, as they provide no insight into the processes that led to the formation of a network in its given shape. Similarly, without studying the content of the communication, we know little about the norms that govern it and the meanings that circulate in it.
Our goal was to conduct an ego-network analysis using Twitter data. This method assumes that each node in the network — in our study, an individual or an organization — is the center of a network of other entities. The term “ego network” indicates that each network is centered around an individual actor. When mapped together, these ego networks can be clustered into “communities”: groups of internally connected nodes. Thus, through network analysis methods, we can classify individual nodes into communities and visualize the network's structure as a community structure.
We began by manually creating an initial list of 200 organizations that meet the definition's criteria. The list was compiled based on Open Future’s database of partner organizations and by reviewing the participants of crucial movement events, such as the Creative Commons summit, Wikimania, or Mozfest events. Each organization in the list was assigned one or more fields of open in which it operates. We intentionally focused only on organizations and not individual activists, seeing them as the core nodes in the open movement.
For each of these organizations, Twitter followers and followed lists were extracted. This way, a list of 211 .410 profiles was obtained, with each profile followed by at least one organization in our initial list. As working with such an extensive network was unfeasible, the analysis was limited to only those profiles followed by a significant number of organizations on our list (threshold of at least 15 followers from the list). As a result, we added 587 profiles to our list.
This group of profiles included both organizations and individuals. We decided to add individuals to our database to acknowledge that the critical nodes of the movement, as made visible on Twitter, are not just organizations but also people. The additional nodes in the sample (both organizations and individuals) were categorized according to the fields of open.
In addition, we added a category called “other” to tag nodes that appear in our network but do not fit into any of the fields of open. Most of these nodes conduct work relevant to the open movement and fit within the broader fields of digital rights and public interest technology activism. This category also includes state actors or international organizations working in digital policy.
Out of over 57.000 relations between the profiles in our list, 18.672 were mutual and included in further analysis. To establish the strength of these mutual relations, we downloaded the 100 most recent tweets per profile to obtain data on interactions, including retweets, mentions, and replies.
This data was used to simulate the network using the D3.js framework and the ObservableHQ platform. A physics-based simulation was used as a reasonable clustering heuristic, and push-pull forces were derived from the analysis of relationships. The strength of the links was determined based on the interaction data from Twitter.
As a result of the simulation, a topological representation of the network of profiles was created, with the simulation creating clusters of nodes that share the strongest relationships. In addition, data on the number of followers (among the nodes in the network) is included to measure the importance of a given node in the network. As a final step, the data on the fields of open was added to the analysis. This allowed us to verify whether the communities identified by network analysis overlap with the fields that we have identified.
More advanced network analysis methods could be used to analyze this data further using quantitative methods. However, the approach adopted was sufficient to conduct this initial analysis, and we believe that further analysis would not necessarily provide additional, actionable data.
One interesting path to explore would be a linguistic analysis of the collected corpus of tweets. This could provide compelling insights into the issues and narratives that organizations and individuals in the movement are communicating about.
Finally, our focus has been on the overall structure of the movement and not the individual ego networks. We are more interested in better understanding the shape of the movement as a network rather than the position and role of individual actors.
An overview of the network visualization follows, illustrated by static versions of the visualized network and its component fields. An interactive version of this analysis is available online as an Observable notebook.
The ego network analysis and the visualization confirm the movement structure that we propose as a system of connected fields of open. The analysis shows clear clusters derived from the data on communication behavior and mutual relationships between the different entities in the movement. These overlap with clusters obtained by categorizing organizations according to the fields of open.
As mentioned, the visualization does not present an objective map of the open movement. The relations and their strength, inferred from Twitter data, are just one aspect of the movement’s internal structure and the relationships between actors. At the same time, Twitter (and other social networks) are an essential space in which this movement exists as a network of connected individuals and organizations. And the communication activities on Twitter express underlying connections and positions of actors in the movement.
A clustered movement. The visualization shows that the open movement consists of several main clusters. These consist of groupings of the major fields of open. Some of the fields of open that we identified previously have not been visible in our analysis — with only a marginal number of actors analyzed being active in the given fields. The main clusters are:
The Open Government / Open Data cluster
The Open Web cluster
The Open Culture / Free Culture / copyright reform cluster
The Open Access / Open Science / Open Research Data cluster
The first two clusters are close, and the third serves as a connecting space between them and the last cluster. This suggests that communication within the movement is focused on these four relatively distinct communication spaces.
Other fields that we identified are less clearly visible as separate clusters. In particular, the field of Free Culture is looser but also serves as a connector to other fields. The distinct nature of this field can explain this as less institutionalized and less connected with a specific domain. Possibly also, the stakes that we identify with this field serve as some form of joint mission for other movement actors. Also, some fields, like Open Design or Open Climate, are not only small but do not form a cluster. This suggests that activity in these emerging fields is conducted alongside work in other fields by actors in the movement.
The Open Source activists and organizations cluster is also distinct in its shape, as it is distributed across the network we analyzed. There are two possible explanations for this. Firstly, this could mean that Open Source is a cross-cutting issue, with organizations and people with Open Source expertise engaging in conversations within different fields of open. Alternatively, this could mean that we are simply seeing a fragment of a more coherent field distinct from this space and therefore, not appearing in our network analysis.
Similarly, a low number of Open Education activists and organizations is surprising, considering the strength of Open Education networks, also on Twitter. Most probably, these actors form a separate cluster that is weakly connected with the other fields of open.
As we mentioned before, Twitter is just one communication space used by activists, and this is particularly relevant for the open-source community, which has championed the use of alternative networks, such as the decentralized Fediverse. A larger segment of actors in this field is probably not using Twitter.
Movement connectors. As in any network, some nodes have a stronger position than others. In our analysis, this can be due to the relevance among other movement actors (measured by the following activity) and the strength of communication activities (measured by retweets, replies, and mentions). We assume that the two factors are interconnected. Our analysis helps identify these critical actors for each cluster.
Our network also identifies movement connectors — nodes that play a role in connecting the different clusters of the movement. These are the key nodes in the space that connect the movement's main clusters. Key movement organizations are located in this space and play this role: Creative Commons, Wikimedia, Open Knowledge Foundation, Internet Archive, and to some extent also Open Science, Berkman Klein Center, and Mozilla. All these movement actors are characterized by work that occurs in different fields of open. In the case of Creative Commons and Open Knowledge Foundation, this also acknowledges the cross-cutting nature of licensing tools they are stewarding.
The role of individual activists. While our initial aim is to map the network of organizations that form the open movement, the network analysis suggests that the movement should be conceptualized as constituted by both institutions and individuals. Naturally, many of them represent organizations, and it is also characteristic of Twitter that communication is often personal. Nevertheless, the analysis results suggest an important role that individuals — especially those functioning as key nodes in the network — fulfill in facilitating conversations and the flow of information.
Open movement and other forms of digital rights activism. Our analysis identified the open movement as connected to other activist networks or digital social movements. An analysis of the nodes shows a cluster of nodes that, while strongly connected with the movement network, do not fit within any fields of open. Instead, the organizations and individuals in this loose cluster can be broadly defined as working on other forms of digital rights activism. This cluster is connected most strongly to the fields of Open Data, Open Government, and Open Internet. It has little connection with Open Science, Open Education, or Open GLAM. The “Other” category also includes actors that connect with the Open Movement network, but fall beyond both the fields of open and other forms of digital rights activism: public figures, public institutions, international organizations, etc.
The network analysis confirms the hypothesis that the movement can be conceptualized as a series of fields of open, related to specific domains, in which the vision of open is applied. The open movement does not have uniformly strong connections between the different fields of open — if these are treated as key structural units. There is a clear split between the two parts of the network. One part focuses on openness in the fields of science, culture, and education — and the other deals with openness related to data, the public sector, and the internet. And it is the latter that is connected with other fields of digital rights activism.
It should be noted here that network connections – or their lack – abstractly represent a broad range of different relationships. A connection might mean strong alignment and/or only communication activity, such as promoting the spread of messaging.
Some fields, especially the major ones, form clusters with strong internal connections. There are also overlaps between some fields (for example, Open Data and Open Government or Open Data and Open Access). This aligns with an intuitive sense of connections between the different fields shared by movement activists.
Some of the connections are less obvious — for example, the strong connection between Open Government and Open Internet activism, coupled with a weaker connection between the latter and other significant fields of open. In our initial analysis of the fields of open, we hypothesized that the field of Open Internet activism might be central and connected to all other fields.
Just as interesting — and important — is the position of the cluster labeled as “other,” which largely corresponds with actors involved in digital rights activism. This is a crucial connection between two related activist networks or movements. Looking from the perspective of the open movement, a digital rights perspective is necessary for the open movement to tackle such challenges as imbalances of power. The fact that a large portion of the open movement is weakly connected to the digital rights network is, in this regard, problematic.
Assuming that the communication network we analyzed represents the open movement, specific actions should be undertaken to maintain and strengthen this movement. The following recommendations are made assuming that such movement coherence is beneficial. The various fields of open can be strengthened not just by activities within a given field but also by those that strengthen the connections with other fields.
Deliberate communication between the fields. Such communication would raise awareness about other fields of open. Clusters in the network show that, in many cases, actors are focused on relations within a given field — yet they can also benefit from a better understanding of what is happening in other fields. This also entails promoting digital rights awareness in the fields of Open Education, Open Science, and Open Culture.
Advocacy as a cross-cutting activity. Our analysis shows copyright reform activism as a network connecting several different open fields. Advocacy, in general, is an activity that benefits from movement-wide engagement. We explore this further in our other report, “Shifting Tides.”
The role of movement connectors. Some nodes in the open movement — both individual activists and organizations — are in a position to connect different parts of the network. These roles are crucial, and as such should be well resourced and supported.
Big organizations and movement solidarity. The movement-wide connections are in many cases sustained by large, established organizations. As such, they should consider themselves stewards of a broader activist ecosystem.
Strategic conversations across the fields. The different fields of open function at a distance from each other. This means that within these fields open activism and other types of activities could have developed in different ways. Learnings made across these fields, and explorations of these parallel developments, can be fruitful in terms of inspiring strategies within these fields, but also for shaping a cross-cutting strategic agenda.