The Prefigurative Politics of Research Software

Research software. We use it everyday, from the most ubiquitous things such as note-taking applications and document viewers to the most sophisticated software packages used in medical simulations or machine learning. So in a rather trivial sense, we all care about it, because we all share a basic need for research software, regardless of our qualitative and quantitative approaches. But what I hope to show in this brief post is that beneath this seemingly functional question lies a fundamental problem, but also, quite remarkably, a global movement addressing it one piece of software at a time.

The silent problem with proprietary research software

There are deeper reasons to care about research software, and more specifically about the type of research software, that we make use of in our work. Our decisions to use a given piece of software over another for our research purposes go beyond practical considerations – despite what many people are made to think based on the marketing strategies and commercial advertising of various software companies. They are tied to crucial political questions of freedom and equality, and feed into a much wider set of social and economic dynamics that involve millions of researchers and students around the world.

But why? What is it about research software that makes it so important to us beyond its technical features and user interfaces? Well, it all comes down to two simple concepts: corporate power and license fees. It is no secret that academics themselves and their universities spend huge amounts of money on proprietary software that have become essential for researchers and students alike. Take the example of Microsoft’s new licensing option for education customers. Their price point seems to be around 35€ for a 6-year plan per user. Now consider the Université libre de Bruxelles (note: this applies to nearly all universities, I just opted for a familiar case), which has some 35,000 students and 5,000 workers (counting both academics and administrators). To keep things simple, if all of them were to start using Microsoft’s software bundle at the same time, this would cost the university no less than €1.4 million.

One might think this represents a rather small amount in the university’s total budget. Sure enough, student fees alone generated €21.2 million in 2022, although the real budget increase for that year was of €4.5 million, which already brings us closer to the Microsoft expense. But here comes the real sting: the ULB also pays for several other software licenses (see the full list here). This includes Labview, that offers one-year subscriptions ranging from 520€ to 3,020€ per user, and SPSS, that IBM charges anywhere from around 3,500€ to over 23,000€ per user. And on top of all this, we also have to factor in the learning management systems that universities deploy to provide students with online educational services (think Moodle). According to a recent estimate, universities around the globe spend a breath-taking $6.56 billion on this particular type of software alone, and this number is forecasted to reach $16.1 billion by 2030. By now, I think, we can all start to see both what is at stake and how much is at stake.

The living alternative: free and open-source software

The real question remains, what can we do about it? This is where we come back to the original issue: our choice of research software. Because we do, in fact, have a choice. There is, and always has been, an alternative to proprietary software: free and open-source software (FOSS). This is not the place to discuss the fascinating history (see Tozzi, 2017) nor the (ambiguous) political principles (see Coleman, 2004) of the FOSS movement. Suffice to say that the existence of FOSS is as ubiquitous as it is unknown: some 85% of mobile phones run on top of the Android Open Source Project, some 96% of the top one million web servers run on Linux (as do every single one of the world’s top 500 supercomputers), some 4.3 billion users visited Wikipedia last year, and an estimated 178 million users run Mozilla Firefox as their web browser.

What matters to us here is the political economy that underpins the FOSS movement (see Weber, 2005; Vanheuverswyn, 2007; Birkinbine, 2020). By definition, FOSS licences make it that a given piece of software is freely distributable and accessible to all. The ‘four freedoms’ outlined by the Free Software Foundation, which are common to nearly all free and open-source software licenses, posit:

• The freedom to run the program as you wish, for any purpose (freedom 0).

• The freedom to study how the program works, and change it so it does your computing as you wish (freedom 1). Access to the source code is a precondition for this.

• The freedom to redistribute copies so you can help others (freedom 2).

• The freedom to distribute copies of your modified versions to others (freedom 3). By doing this you can give the whole community a chance to benefit from your changes. Access to the source code is a precondition for this.

These legally-enforceable freedoms prevent most if not all imaginable restrictions on any software under a FOSS licence, hence software that is free (as in freedom) and open-source is nearly always, barring exceptions, available for free (as in free of charge). While this radically breaks with the conventional logics of commodification under capitalism, a number of successful business models have been built around free and open-source software, with some FOSS companies such as RedHat, Github, and MongoDB reaching financial market valuations in the billions of dollars.

Millions of people around the world collaborate on countless FOSS projects, from the smallest hobby of a college student to the largest software commons that receive considerable financial backing from large corporations and governments (such as Gitlab, the Tor Network, the Linux kernel, Chromium, the Free Code Camp, and VistA). While most FOSS projects are developed by volunteers and funded by individual donations, some also benefit from the paid contributions of employees working either for small, medium, and sometimes (very) large tech companies (including Google, Meta, IBM, and even Microsoft), or for government administrations and universities. The FOSS movement, in a nutshell, is a complex, somewhat conflicted, but above all massive community of programmers, contributors, and active users of truly global proportions that has completely transformed the production, circulation, and appropriation of computer software forever.

It is hard to understate the positive impact of the FOSS movement on global inequalities under capitalism. All around the world, and especially in the Global South, government administrations, schools, hospitals, non-profits, small businesses, and individual households have made use of countless software technologies distributed under FOSS licenses whose proprietary counterparts would have proven financially onerous if not completely inaccessible altogether (see Weerawarana and Weeratunga, 2004; Barry, 2009; Sowe, Parayil, and Sunami, 2012; Aminpour, Sadoughi, and Ahmadi, 2013; Johnston et al., 2013; Sahay, 2019; Thapa, 2020; Sandberg, 2023). Indeed, an inestimable amount public resources have been diverted from private profits in developing countries with measurably positive social and economic effects (see Papin-Ramcharan, 2007; De’, 2015; Silva, Coutinho, and Costa, 2023). Unfortunately, many people and organisations find themselves dependent on proprietary software due to their various lock-in effects. Two main issues here are the closed-sourced technological standards on which proprietary software tend to operate (i.e. Microsoft pushing the ‘.docx’ document format on users instead of the inter-operable ‘.odt’ format), and the costs of retraining personnel on specialised software. Nevertheless, transitions to FOSS alternatives have proven possible even when these obstacles incur, as the case of the Italian Ministry of Defence switching from Microsoft Office to LibreOffice illustrates.

Research software: the choice, and change, is in our hands

The same is true for universities. Were most if not all of the research software used by academics, students, and administrators free and open-source, this would create massive savings in public expenditures that could serve to fund research, cut tuition fees, hire more people, and renew infrastructure (for an estimate of savings incurred by switching to FOSS equivalents in scientific research, see Pearcke, 2020). Think about it: if all universities in Europe, let alone Belgium, were to pool their resources in order to develop and maintain research software under free and open-source licenses available to all at no financial cost, this would be nothing short of socially transformative. All researchers would benefit equally from high-quality software based on interoperable, open standards, and all students would benefit from the massive financial savings of universities on licence fees. At the end of the day, making research software freely accessible through FOSS licences has the same effect as making scientific knowledge freely accessible through open access publications: it benefits everyone in society.

The beautiful thing is, several universities around the world are already doing just that. In fact, universities have been involved in the free and open-source software movement from the very beginning. The first academic organisation that got seriously involved in the FOSS movement was the University of California, Berkeley, where the Computer Systems Research Group (CSRG) developed the Berkeley Distribution System (BDS) in 1978 under the BDS software license – a FOSS license that remains widely used to this day. Another academic pioneer in the free and open-source movement worth mentioning is the Massachusetts Institute of Technology, where researchers came up with another popular FOSS license at the end of the 1980s, the MIT license.

Ever since then, many academics, research labs, and universities have been involved in free and open-source software projects. Some of the most prominent examples include the main programming languages used in quantitative research, Python and R. The former was started by Guido van Rossum from the Centrum Wiskunde & Informatica of the Dutch Research Council in 1991, and the latter by Professors Ross Ihaka and Robert Gentleman from the University of Auckland in 1993. One of my personal favourites, Orange, which allows researchers to conduct data analysis visually without any programming skills required, was likewise started by researchers at the University of Ljubljana and the Jožef Stefan Institute in 1996.

Professor Isaac I. Ullah (San Diego State University) maintains a very comprehensive list of FOSS software packages for various use cases in academic research. There are a couple types of software packages he does not (yet) mention, however, that I think are worth pointing out. First of all, when it comes to searching for the most relevant scientific papers (‘literature mapping’) using citation networks, four fantastic FOSS alternatives to proprietary equivalents are Inciteful, Citation Gecko, OpenKnowledge and ResearchRabbit (as listed by Ujjal Marjit). Secondly, when it comes to real-time online collaboration software, there are a number of FOSS alternatives to Google documents such as Framapad (for general purpose) and Cryptpad (for sensitive data). Thirdly, when it comes to writing down and structuring our endless stream of thoughts, Zettlr stands out as one of the very best free and open source alternatives to proprietary software such as OneNote. Finally, being a computational archaeologist, Isaac Ullag does not mention any Computer-Assisted Qualitative Data Analysis (CAQDA) software packages. I happen to be a more qualitative researcher, and would thus like to mention in-depth three relatively unknown yet robust FOSS alternatives to proprietary equivalents such as NVivo and Atlas.ti. All of them were born out of beautiful and unique collaborations among researchers and active users, and can meet most if not all of our needs as qualitative researchers when dealing with textual data:

• Catma, created by scholars in the digital humanities along with IT experts and visual designers at the Technische Universität Darmstadt around 2009, with the financial support of multiple German academic institutions. It runs straight from the web browser, offers a sleek, modern, and intuitive interface, and is perfect for collective research projects (i.e. sharing code books, reviewing the coding of peers, etc.). The website is very complete in terms of documentation and tutorials, and numerous academic articles (both empirical and methodological) have been written on it.

• QualCoder, originally created by Professor Colin Curtain, a pharmacologist, at the University of Tasmania in January 2019, on his own spare time with the help of volunteers and individual donations. Based on PyQDA and RQDA, it runs locally on all major operating systems. While its user interface is rather minimalistic, it offers more features for data visualisation (word clouds, charts, graphs) and semi-automated coding than the other two programs. There are some great online tutorials (here and here), solid documentation can be found on the Github page (Wiki section), and there are interesting methodological articles on its deployment (see Brailas, Tragou, and Papachristopoulos, 2023).

• Requal, created by a group of sociologists at Charles University and the Czech Academy of Sciences, with funding by the Technology Agency of the Czech Republic, in 2023. While it is the newest kid on the block, it seems to offer a promising synthesis of Catma’s user-friendliness and online collaboration features, and QualCoder’s feature-completeness and local deployment capabilities (which matters for those with sensitive data). The documentation on the Github page (Wiki section) is very clear and comprehensive, and they have also published twelve mini-tutorials detailing the features of Requal (accessible here).

Of course, proprietary software packages will often have some niche features or refined look to them that FOSS equivalents do not always offer. However, it would be both unfair and unrealistic to expect a handful of people, working part-time or even in their spare time on these pieces of software with (very) limited resources, to match all the functionalities and the savvy ‘material design’ of proprietary software which is being sold at a hefty price by specialised companies. But the crucial point I hope to convey in this post is that beyond these practical and aesthetic considerations, we are faced with a real political decision as users of research software. We can either i) perpetuate the status quo, that is, ignore all the public resources that are being wasted on exorbitant license fees, at the expense of say more research funds or better working/learning conditions, or we can ii) adopt and promote FOSS software in our research communities, and thus contribute to a net positive change for both our universities and society at large, because FOSS is always, by definition, to free and open for all.

Truth be told, almost all software being used by universities today have a credible FOSS alternative. And to be clear, I am not just talking about the specific research software students and academics use for their empirical analysis – I mean office suites, email providers, cloud storage, video conferencing, and online learning platforms. So by deciding to promote FOSS solutions to our needs as researchers and students, we are contributing to a much broader movement for change within our universities. And if, through our principled software usages and our campaigns for change, we manage to turn the tide so that more universities begin to pool their resources into FOSS projects, then we will have played our part in the prefigurative politics of a global digital commons.

By Édouard Hargrove (picture from lorilorilo on Pixabay).