Titus Stahl

Why I refuse to use non-free software

(licensed under CC-BY 4.0, please let me know about factual errors and spelling mistakes.)

As an academic, my primary activities, next to reading and listening to papers and lectures, are writing, teaching and participating in the administration of my department as well as the collective administration of several journals.

Some of these activities are essentially collaborative (such as teaching or editing journals), others are sometimes so (writing papers). In the course of my collaborating with other colleagues, they often ask me to use programs such as Skype, Dropbox or Microsoft Word. I then refuse to do so, on the ground that these programs are not Free Software. This argument is often met with the attitude “in principle you are right, but please make an exception for this case in which there is an obvious practical benefit”.

As there is often neither time nor opportunity to engage in a lengthy discussion about this question, I will here briefly lay out what the problem is and why I think that computer users in general and academics in particular should refuse to use non-free software.

Free Software

Free Software is often confused with software that is free of charge (“gratis software”) and with the more related type of “open source” software. It is therefore worthwhile to define these concepts clearly.

The main distinction between Free Software and other software is a question of which rights you have in regard to the software program in question. In most legal systems, the creator of a software program holds exclusive rights in her or his product that allow him or her to exclude other people from using, modifying or distributing it. This holds true unless the creator transfers these rights to these other people.

Therefore, most software which is distributed comes with a license that specifies what rights the user of the software has.

A piece of software is Free Software if and only if its license extends the following four rights perpetually to its users:

  1. the right to use the software for whatever purpose
  2. the right to study the program and to adapt it to one’s own purposes
  3. the right to share the program with others
  4. the right to improve the program and to share the improvements with the public

These four rights are all individually necessary and jointly sufficient to give a user of a software product control over that software product. “Control” means here that you can use, modify the program and communicate about it in any way in which you see fit.

This translates into control over your computer: When you buy a computer without any software on it, you may execute all possible instructions on that computer, change every bit of data in its memory and communicate with others about what your computer does without restriction.

What happens when you then buy a proprietary product (such as Microsoft Windows), agree to its license and install it on your computer? Then, there are certain instructions you may no longer execute (such as modifying or manipulating the software which is prohibited by its license terms); you may not change its representation on your hard disk and you may certainly not communicate any part of this representation to others (i.e. copy the software or parts). In addition, because you may not execute certain instructions (such as controlling the program with a debugger), you are effectively forbidden to know what your computer does at any given point in time. In short, you lose much of the control over your computer you had when you started.

Using non-free software thus directly translates into a loss of control over your computing equipment.

It is irrelevant for the question of whether something is Free Software if the rights specified above were transferred to you free of charge or in exchange for a payment. Free Software can be sold, although the fact that each person to whom it is sold can sell it (or give it away) themselves, quickly creates competition for the original creator, so that Free Software tends to be free of charge.

The four rights, in particular the second and fourth, often practically require that users have access to the source code (that is, the human-readable text, written in a programming language) of a program to be able to exercise these rights. Thus access to the source code is (almost always) an implication of software being Free Software. However, in contrast to what the term “Open Source” suggests, having such access is not sufficient for a program to be Free Software. Many software companies allow important clients to look at the source code of their programs, for example, to verify that there are no security holes, without granting them any of the rights which would make the program Free Software.

Free Software and Political Freedom

Why should one refuse to use any software that is not Free Software? There is an answer which should be especially clear to political philosophers.

My argument is the following:

  1. A large part of our communication with fellow human beings depends, already today, on this communication being processed by computers (and it is virtually guaranteed that this share will increase in the future).

    We depend not only on computers processing the communication when we write e-mails, download books or web pages or when we use internet telephony. Even when we use the “ordinary” telephone network, computers process our calls.

  2. All contemporary practices of democratic self-rule, most forms of political participation and an increasingly large part of those practices constitutive of private associations (families, friendships, etc.) on which our personal self-realization depends are now practices involving digital communication. In fact, this is now practically unavoidable, given the digitization of much of our communicative infrastructure.

  3. The software running on the computers which process these communications shapes these communications. Computers allow for the representation of data in certain formats, their algorithms control who can access the data and who cannot (shaping both the scope of participation and surveillance), and they allow for certain communicative acts and disallow others.

    This leads to the following situation: If our communication with your friends, our political allies or our elected representatives, the practices in which the public deliberates and the data processing which is part of our personal lives (involving our e-mail accounts, our profiles on social networks, our phone book) are shaped and mediated by computers, anyone who can effectively control these computers, can control the way in which we shape our own lives.

  4. Computers that run non-free software are not controlled by the user of that computer but by the creator of the software.

    In principle, computers are designed to do whatever they users want them to do. However, if you execute a piece of software on them of which you do not know how it works, and where you have accepted that you have no right to change how it works, you hand over the control of your computer to someone else, the creator of the software.

  5. Being autonomous, i.e. deciding yourself about the shape of your personal, social and political life (either as an individual or together with others, collectively) is therefore impossible when the communications technologies that enable such deciding are under the control of others; that is, when you use non-free software.

    It is important to emphasize that the notion of liberty here at stake is not the liberty of “non-interference”. I do not want to claim that the producers of non-free software today regularly interfere with the way in which people want to communicate or to make collective decisions (although this, of course, occasionally happens). Rather, those who own these rights have always the power to do so, whether they use this power or not. In other words, if you use proprietary, non-free software, you are at the mercy of these other people not abusing their power. When you use Free Software, they do not have such power over you (and everyone else), because you can yourself study and determine how the essential personal and social practices of the digital age are shaped by the software that enables them.

A practical application of this argument is, of course, surveillance: If you use non-free software to communicate, you are unable to determine whether this software protects your data and whether it has any security holes that allow others to spy on you. You can only study the software and fix such holes (or pay third parties to do so), if you have the four rights constitutive of free software. As the NSA affair has shown, this is by no means a merely theoretical, but a highly practical threat.

The iPhone as an especially egregious example of unfreedom

To cite just one extremely disturbing example, many of my acquaintances use Apple’s phone product. This is especially worrying, since the iPhone not only runs a non-free operating system, but Apple actually reserves the right to decide what software you are allowed to run on your phone. If you do not have a developer license, you cannot buy software for your phone on the free market, but only via Apple’s store. Apple regularly rejects software from being published in this store, for a number of reasons.

To name just one example, Apple has rejected an app that tracks U.S. drone strikes for the reason that the audience may find the content “objectionable”. Thus, people who have paid for a small computer and who want to run some piece of software on it are forbidden by Apple to run that software — which is unquestionably not illegal — because Apple worries about other people’s opinions.

Even more worrying is the fact that you can only publish software in the App Store if you agree to restrict your users’ freedom, so that it is impossible to publish Free Software in the App Store, in effect making it impossible for iPhone users to acquire any ethically acceptable software.

Free Software and Academic Practice

Next to the general reasons against using non-free software, there are particular reasons that apply to academics which make it even more important for them to use only morally acceptable, Free Software.

  • Academics deal with information. Academic practice is the gathering, transmission and publication of information, either to colleagues, the public or to students. Traditionally, universities have prided themselves of being independent from external pressure. In the past, this was often easy to achieve, especially in the humanities, because the means of production were relatively cheap to acquire, compared to industry. If academics use tools for their practice in regard they are not free to use them and to understand them as they would like, this is a grave danger to academic freedom. Academics cannot accept a system which even contains the option of commercial companies deciding about which information may be stored or transmitted, nor can they accept a system which enables the possibility that their research is subject to surreptitious surveillance.
  • Academics is traditionally a field in which nor private property in information, but sharing of information is celebrated and part of the professional ethos. Non-free software is incompatible with this ethos.
  • Academics transmit not only knowledge to their students, but also train them in using the scientific method. While there is often a separation made between the “really important methods” and “mere tools” such as word processors, this is not a distinction in reality and it will become even less so, the more sophisticated our software tools become. Many universities promote, for example, reference management systems which are non-free. When students leave the university, they lose access to a “campus license” covering this software and, thereby, to their own references.
  • Academics have a duty of care towards their students. They require their students to experiment with theory and with arguments and thus require them to engage in a practice which is only free when it is not subject to surveillance — most notably in fields that concern politics or ethics. If academics require or promote the use of non-free software, they might actively harm their students. Finally, non-free software is often expensive and using it excludes students from poorer backgrounds.

What are our moral duties?

As academics, we have, on my view, the following moral obligations:

  • We never ought to require or encourage either colleagues or students to use non-free software.
  • We should, as far as possible, use exclusively Free Software ourselves. As software use always produces a “network effect” (when a lot of people use software X, it will become more useful to other people, due to easier sharing of information in the same format etc.), no use of software is neutral.
  • We ought to educate our students and peers about Free Software
  • We ought to reject all attempts, either by colleagues or university administrations, to force faculty to use non-free software or to use university resources to promote such software.

Some of these duties are absolute (e.g. the duty not to require others to use non-free software), whereas some are defeasible. The duty to use only Free Software ourselves can legitimately be violated when research results cannot be achieved any other way or when one’s career will suffer in non-trivial respects if one refused to do so (e.g. if a funding agency or an important journal requires such use). However, if there are such cases, we ought to note our disapproval of this being so.

Free Software Alternatives

Most of the non-free software that people in philosophy use can easily be replaced by free alternatives

  • Instead of Microsoft Office, use LibreOffice.
  • Instead of Endnote, Citavi, Mendeley or other non-free reference management programs, use Zotero.
  • Instead of Dropbox, use OwnCloud.
  • Instead of Skype, use Jitsi or Firefox’s inbuilt video chat. (Update November 2016: I now recommend Riot for audio and video chat).
  • Instead of Microsoft Windows or MacOS, use a GNU/Linux variant such as Ubuntu or Debian.

For more alternatives, see also Prism Break.

Common Objections

When I make these arguments, I am usually met with some objections that concede the general point but want to argue that, in this particular case, I should accommodate the request and use non-free software. While, as noted above, the obligation to use Free Software is sometimes defeated, most often it is not. I would like to mention two such arguments here.

“But Free Software is missing important functionality X”

Often, proprietary software has some features that are missing from a Free Software equivalent. This is, of course, unfortunate. Before you ask me to use non-free software, please consider the following questions:

  • Is the feature merely convenient or essential to reach an objective? Mere convenience obviously cannot defeat moral obligations.
  • Is the indispensability of the feature a result of your unfamiliarity with Free Software? Many features of Microsoft Word could be replaced by a work flow using plain text files and text-based command line tools. Many people do not know how to use these tools. It’s a lot of work to learn how to use them, but this is to be preferred over the alternative of using non-free software. It may be acceptable for me to use non-free software while you are engaged in learning to use alternatives. But if you refuse to do so, it is not me who is uncooperative, but you.
  • Have you taken every effort to improve the Free Software alternative? Did you check whether the missing feature is already registered in the Free Software project’s bug tracker? Did you donate money for its implementation on BountySource or FreedomSponsors? If not, consider doing so instead of giving up. Many people pay directly or indirectly for proprietary software but refuse to pay a small amount of money to Free Software projects which would provide them not only with the same features but also respect their freedom. This is not a consistent moral position.

If the fact that the functionality is missing in Free Software is either unimportant or simply due to the fact that proprietary software is taken to be the standard approach, there is no reason to give up on Free Software.

Can’t you make an exception just for this time?

People often say: “You are right, but now I am busy and can’t be bothered. Why are you so inflexible to not make an exception for this time?” The answer is simple: Because every time that I make an exception, I contribute to the problem’s further existence. Proprietary software is supported by network effects which can only be avoided if people learn to use other software. This is always inconvenient.

Network effects just mean that the dominant alternative is always more convenient. If one takes convenience as a conclusive reason in the decision what to do, the dominant alternative will never be replaced. Thus, there is a collective action problem when morality conflicts with convenience: Everyone would like to do the moral thing, but only if it is convenient, which requires that all others change their behavior first.

This collective action problem can only be solved if some people refuse to use the dominant alternative, at which point it becomes more convenient for everyone else to use the moral option. This is what I am doing.

Doing the right thing will pay off, morally and instrumentally, in the long run. If there is no specific reason beyond your inconvenience, making an exception this time means I would — if I am consistent — have to make it always. I do not want to do this.