Skip to ContentSkip to Navigation
Jantina Tammes School of Digital Society, Technology and AI JTS Themes A Roadmap to Data Autonomy
Header image Data Autonomy

Data Autonomy and Academic Freedom

Date:14 February 2024
Author:Titus Stahl


As universities become more and more dependent on digital infrastructure provided by Big Tech companies, many researchers are increasingly worried about the resulting phenomenon of “digital dependence”. There is the worry that it will be too costly to switch to alternatives, once universities have moved to a specific platform. There are concerns about data privacy, and the protection of confidential information. While these concerns are all legitimate, I argue that there is also another issue which directly relates to one of the core values of a university: academic freedom.

It may seem as going too far to say that the reliance of a university on the Google or Microsoft platforms endangers the academic freedom of its staff and students. After all, there is nothing about these products that keeps academics from pursuing their research interests or students from exploring whatever topic they like to study. Not only do these platforms typically not impose any limits on the content of documents that they can be used to process (except, of course, for the case of illegal content), we can also assume that the companies that run them are willing to agree that they will not censor any academic content. If this is true, is it not enough to safeguard academic freedom?

I will argue that it is not, and that universities’ commitment to academic freedom requires them to take more control of their data infrastructures than they have achieved so far.

What is Academic Freedom?

To understand why this is the case, we need a more precise definition of what academic freedom means. Intuitively, one might think that one enjoys academic freedom whenever one can explore any kind of subject and defend any kind of hypothesis that one wishes to. However, this may not be enough.

Imagine, for example, the case of a researcher living under an authoritarian regime. Her government is happy to interfere with critical researchers whenever it wants, and often does so. For the sake of the argument, assume that she studies underwater basket weaving. As it happens, the current cohort of government bureaucrats has no interest in underwater basket weaving. As a result, she faces no interference with her research. Nevertheless, I would argue that she does not enjoy academic freedom in a full sense, since the government still could, at any point and on a whim, interfere with her research.

Similar to what so-called “neorepublicans” in political philosophy argue regarding freedom in general, there are therefore strong reasons to say that, to enjoy academic freedom, it is not enough that one does not, in fact, face interference, one needs to enjoy robust protections against possible interference (by the government or others). Academics need to have the confidence that it is not merely a matter of luck that they can explore whatever questions they think are interesting, they have to know that there are rules that protect them.

However, academic freedom is not unlimited. It does not extend to researchers making fraudulent claims, to instructors intentionally misrepresenting the facts, or to students’ plagiarism. Academic freedom means the robust protection of research, teaching and study within the bounds of the rules of academic integrity and the limits imposed by the obligation to fairly evaluate the evidence.

The question then becomes who decides where these limits are. My claim is that an academic community does not enjoy academic freedom when others impose their interpretation of these limits on it - for example, if politicians begin to make decisions about what is acceptable evidence for a scientific claim. Academic freedom implies the robust, institutionally protected capacity of an academic community to set its own standards, and a robust, institutionally protected right of individuals to pursue their academic projects within the limits of these standards.

This does not mean that academic communities cannot collectively go wrong. In the past, some scientific insights were unfairly dismissed because they violated the narrow ideas of the academic community of the time about what constitutes proper scientific evidence. Collective academic freedom can also be misused, and there are always legitimate controversies about how academic communities should govern themselves. However, to defend academic freedom means to say that these disputes should themselves ultimately be resolved by open, academic debate in a self-determining community of researchers, teachers, and students, and not by outside intervention.

Data Dependence, Infrastructures of Communication, and Robust Academic Freedom

What does this have to do with the reliance of universities on Big Tech? To understand this, I want to introduce another dimension of academic freedom - the freedom to communicate and to shape one’s communicative environment.

The core contribution of universities to society is two-fold: On the one hand, universities teach specialized knowledge and skills that enable citizens to contribute in specific ways to society as participants in public debate, as professionals, and as researchers. On the other hand, universities conduct scientific research, the results of which then allow for a better understanding of the social and natural world, for new forms of technology, but also for more informed societal decision-making. Both contributions depend on communication. Progress in scientific knowledge is only possible if researchers coordinate their activities, discuss and debate their results, and communicate them to the public. Teaching, even more obviously, is an essentially communicative process.

When universities depend on data infrastructures that are provided by Big Tech, they do not only get a more efficient version of the communication tools they always had anyway. Google Docs is much more than a more efficient typewriter, since it allows for new forms of collaborative editing and commenting that are unique to that platform. Social media platforms are not just better versions of the newspaper or the post office, since they allow for new forms of interaction (liking, re-sharing, but also social media harassment). Research Information Management Systems, like the UG’s Pure, are not just a fancy way to put your CV online, since they allow for a calculation of productivity statistics and for new ways to measure academic performance.

In all these cases, universities currently enjoy no real power over the design of tools that are essential for the processes of communication that allows their members to contribute to their core mission, although this design fundamentally shapes the way in which they communicate. In particular, the design of these tools shapes how they interact, it normalizes certain kinds of information flows, and generates some expectations of privacy, and undermines others.

For example, we may very well think that a Learning Management System such as Brightspace would enable more desirable forms of communication between students and instructors if it had less monitoring capacities (such as instructors being able to see in detail which student has accessed which subpage at which time), and instead provided more tools for argumentative deliberation than its relatively clunky discussion boards. Similarly, a research information system like Pure might better serve the purposes of academic cooperation if it had better tools to find and connect to people who you might want to cooperate with, rather than generating predictive “research intelligence”, or enabling meaningless competitive rankings. However, as scholars whose work feeds Pure, we do not meaningfully control its design. That design is shaped by Elsevier’s corporate interests rather than by academic values.

But even if we assume that, presently, these tools are perfect for our use, once our data has been moved onto these platforms, moving it out again will be costly, and therefore it is likely that we will have to go along with many of the future design decisions of Big Tech.

While these platforms currently impose few limits on what content we can use them to discuss, depending on them leads to a situation in which the academic community lacks robust control over its means of communication. If we assume that, to enjoy full academic freedom, an academic community need not only be protected regarding its capacity to set the rules for academic behavior, but also in its capacity to make decisions about how to cooperate and how to communicate, this lack of control undermines the very collective basis of academic freedom on which the academic freedom of individual researchers, teachers, and students depends.

While it is, of course, always possible that we get lucky, and the way in which Big Tech designs its platforms happens to conform to our own judgment of how we ought to communicate, our present dependence on Big Tech platforms will always make this a matter of luck. As a result, our academic freedom does not enjoy robust institutional protection. Similarly, even though it is not a university-mandated platform, the example of X/Twitter (on which much academic communication was starting to depend, until recently) demonstrates how quickly one’s luck can run out. In such cases, leaving a platform that everyone has bought into is costly. To ensure robust protections of academic freedom, universities need to be in control of the design of their communicative infrastructure.

There are some glimmers of hope, however. One of them is SURF’s inspiring attempt to create a cooperative social media platform for Dutch scholars and students on Mastodon that aims to allow academic communities to control their communicative infrastructure. Another one is the University of Groningen Press that allows scholars to run their own journals on, and contributes to, the open-source OJS platform.

Universities ought to build on these examples and cooperate in creating, funding, and running their communication infrastructures, such as scholarly governed publishing and social media platforms, themselves, based on openly available and freely adaptable Free Software. This way, they could not only avoid the many problems associated with subjecting their students’ and employees’ data to the legal regimes of foreign governments. By not being subjected to the profit imperative of Big Tech, they might even save money. But most importantly, they could create an infrastructure that protects the freedom of the academic community to shape its own modes of communication.

About the author

Titus Stahl
Titus Stahl

Dr. Titus Stahl is Associate Professor of Philosophy at the Department of Ethics, Social and Political Philosophy at the Faculty of Philosophy. He does research on theories of power, the methods of social critique, the political philosophy of surveillance, and the theory of the public sphere. His interest in data autonomy stems from the belief that democracy requires a public sphere that is not subject to the power of a few, large corporations, and that academic freedom is an essential value of a democratic society.