Persuasion 1

(This is the first of two articles in response to Blaise’s question in the comments below. It is intended as part of a conversation.)


In this blog post I will consider why it might be that a musician should licence their work as BY-SA rather than NC-SA, giving up royalties from public performance of their work as a result, rather than accepting Stallman’s views on the requirements of freedom for works of expression. I will argue that an ethic of freedom of speech rather than freedom of use applies to cultural works, and that the economic harm this may cause for current business models is both acceptable as a moral consequence and can be offset by business models that are already being proven.

Free Software

Stallman’s conception of Free Software is the freedom of individuals to utilise the functionality of software as they see fit. I will call this “freedom of use”. The ethical value of freedom of use is that it supports the freedom of individuals to pursue their own ends within society in a pluralistic way.

It is not unusual for people to complain that they wish to be free to exploit the exchange value of software (including software that they themselves have written) rather than its use value, and that having to allow other people to be able to use the software without restriction affects this freedom. But this is not freedom of use, and indeed is not freedom in general as it involves reducing the freedom of others.

Stallman’s model for software freedom was the 1970s MIT AI Lab. The social contract of the AI Lab embodied the principles and benefits of free use and development of software. It is important to recognise that this is the historical emergence of an (approximately) ideal model rather than a historical accident or contingency. It is also important to recognise that Free Software is reform with a definite model in mind rather than radicalism with an unknown trajectory.

Non-Software Freedom

Stallman has written about the concept of freedom for non-software work. He divides such work into reference, opinion and expression, with different freedoms required for each. Reference material, like software, is a tool to be used to realise your own ends and so should be free. Opinion is designed to have some effect within society and so should not be misrepresented. It should therefore be freely distributable but not modifiable. Expression is not used to achieve or inform anyone’s ends within society and therefore need not be freely copyable. It is also a product of someone’s personality so like opinion it need not be modifiable.

The categories that Stallman describes are guided by the principle of freedom of use. As the usefulness of each category as a means for achieving pluralistic ends reduces, so the requirement of freedom reduces.

I would also argue that there is not such a neat break down of works into separate categories. To take the example of music, a song may be opinion (a satirical song), reference (schooling or tribal songs), expression (punk) or entertainment (corporate punk). Or it may a combination (a protest song may be opinion and expression for example).

Free Culture

As freedom of use is the basic freedom for software, freedom of speech is the basic freedom for culture and cultural works. Supporting and enabling freedom of speech, and opposing censorship, is the pursuit of cultural freedom. Speech here includes art, music, film, and other non-verbal and non-textual forms. Authorship therefore includes artistic, musical, dramatic, cinematographic and other forms.

An author may need to use other speech through reference, parody, satire or quotation (of which sampling is a form) in order to achieve their ends. This is true whichever of Stallman’s cultural categories the work they are producing and refering to fall in. In expression or entertainment, freedom of use peters out (unless, presumably, those works are represented as software) but freedom of speech remains.

The ethical value of freedom of speech is that, like freedom of use, it is vital for the freedom of individuals to pursue their own ends within a pluralistic society. There are many historical models of robust cultural exchange and derivation, from Homeric Verse through Shakespeare and political pamphleteering to Blues, Jazz and dance music and contemporary documentaries and satire.


As with software, it is not unusual for people to complain that they wish to be free to exploit the exchange value of cultural works (including works that they themselves have written) rather than its use value, and that having to allow other people to be able to use the work without restriction affects this freedom. But this is not freedom of use, and indeed is not freedom in general as it involves reducing the freedom of others.

Software freedom requires that free exchange not be forbidden. There is nothing wrong with charging for software, but it is wrong to forbid people from exchanging software that they have paid for. This is supported by the traditional economics of software production, where large institutions pay for the production of software as an incidental by-product of their main operations. Software is a cost rather than a revenue stream in this scenario, and Free Software reduces that cost.

The economics of culture are similarly institutional, with artists and singer-songwriters effectively consultants and freelancers for galleries and record companies. Unlike software ,“content” as culture is regarded by the culture industry is the core of their business rather than an incidental by-product. The production of content is a cost, to be sure, but its distribution is the only revenue stream and free exchange has the potential to undermine this.


To prevent such loss (for individual producers as well as institutional distributors), Creative Commons have produced a “NonCommercial” licence that reproduces the freedoms guaranteed by Free Software “copyeft” licences such as the GNU GPL but with the proviso that use and distribution of the work must not be “commercial” in nature. This is the “Creative Commons NonCommercial-Attribution-ShareAlike License”, or NC-SA for short.

It is not clear that NC-SA achieves its aims. The licence explicitly allows p2p sharing, which is classified as commercial activity under US law and is the development most blamed for loss of revenue by the recording industry. NC-SA allows people to download work over ISP connections using a personal computer and store it on an MP3 player or burn it to a CD then print it out at copy shops. These are all paid goods and services, making money from an individual’s use of an unpaid-for NC-SA work. The only person who doesn’t make money in this scenario is the person who has released the work as NC-SA.

Having given away the ability to freely copy work, the major loss of revenue according to the recording industry, NC-SA tries to lock the stable door not on the author losing money but on other people making money, and fails to do so except in the most direct cases of exploitation.


NC does not stop people making CDs of work, incorporating it into their videos, incorporating it into promotions for events or putting it onto peer-to-peer networks. These are all uses an author could expect to be paid for. Rather than ensuring that the author is paid, NC tries to prevent anyone else from being paid. It fails in all but a few cases. It makes the work into a white elephant, something that can only cost money. Or it makes the work free advertising, paid for by the consumer.

Copyleft, Attribution-ShareAlike or BY-SA (without the NC) in Creative Commons terms, discourages commercial exploitation of work in film, video, and o
ther derivative work. Where it does not, the resulting work will be Free, which is an ethical and reputational gain. BY-SA doesn’t prevent commercial redistribution, but then NC doesn’t prevent noncommercial distribution, which is the main competitor for authors. The main sources of actual revenue that BY-SA doesn’t manage that NC-SA does are commercial distribution of work in CD form and on commercial radio.

It is possible to compete with others copying your work, turning their efforts into promotion. iTunes Music store does. Radiohead’s “In Rainbows” release is an example of comprehensively and successfully competing with leakers and filesharers. The Happy Mondays printed their own bootleg t-shirts and got a friend to sell them outside venues to compete with the real bootleggers.


NC prevents people selling CDs or downloads of work and playing it on commercial radio without paying performance dues. Under SA, the former can be competed with successfully through good release management. That leaves performance rights. Again, commercial radio can be competed with via streaming. And it has positive reputational network effects (it’s free advertising, just make sure you have something to actually sell or be paid for).

Hackers do not complain when a distro packages their work and charges for distributing it. It is recognised that the distributor is providing a service. They provide new users and developers for projects, and build the reputation of projects that they feature. Over time some have come to invest back into the production of the software that they have benefited from.

Authors can view commercial distribution of their work similarly, even where, like distribution of software that one has written, to charge would make money where distro packaging does not. But unlike hackers they can compete with and benefit from this directly. A song can be performed, packaged in a deluxe format or dedicated in a way that a JPEG decompression library cannot, although the distribution of both can lead to further work.


Distribution allows people to hear speech. Where changes to that distribution cause economic harm that reduces authors ability to speak, this can be addressed. For example by reputation economics of scale to offset the loss of performance royalties through network effects on the things you do charge for. That is better than having to pay MTV to play your videos, which is what they wanted. And it will allow ways of your fans finding you to develop that would otherwise be stifled, such as Internet radio.

The current economic ecosytem for music is very complex, with risk and rewards shifted between participants over time and investment and payment flowing to and from many different avenues. This system worked well enough before the Internet but is currently broken and to restore it would cause far greater social harm than just the current lawsuits.

One possible replacement would be similar to the business model of Radiohead’s “In Rainbows”, but with feeling. Release paid downloads before anyone can leak the masters, follow up with deluxe vinyl and digipacks, see who will pay to wrap video sessions with advertising, do sell CDs, perform live as much as you can, and make sure you sell lots of merchandising. This isn’t the only way of doing it, but it is a drop-in replacement for the current model.


Free Culture is an ethical matter. As with Free Software, economic concerns are secondary. Stallman’s categories do not capture this ethics. NC-SA does not protect it, or very much else. Lessig describes a perfect storm of technology, law and economics threatening free speech. Free speech must be protected as a vital part of open, pluralistic society. Any harmful economic effects of this can be borne *if* you find the ethical claims of Free Culture convincing and its aims worthwhile.

Posted in Free Culture
4 comments on “Persuasion 1
  1. Rob, you are my new hero.
    Thank you for such a clear and coherent response, I anxiously await the second. I am already convinced, but if you’ve more to say then it can only help to increase my understanding, resolve and ability to explain my decision to others.
    As an aside, Stallman’s position on works of expression seems to have changed over time though. When I heard him speak last summer, he did advocate for non-commercial sharing, both public and private, and seemed to be opening up to (but still undecided on) the idea of the derivative. That doesn’t change your argument though.
    To play devil’s advocate a bit more though, the non-commercial provision doesn’t prevent competition with commercial stream, but then some will argue that certain levies or royalties ought to compensate artists in light of this, like the Songwrtiers’ Association of Canada’s proposing to add a $5/month license fee to ISP charges. I’ve heard some convincing criticisms of proposals of this sort, but I’d be interested in hearing your thoughts if you have a chance.

  2. Rob Myers says:

    Oh wow. Thank you for your kind words. Now just make sure you get rich off t-shirt sales otherwise I will feel soooo guilty. 😉
    I know that Stallman has kept thinking about cultural works. I think he said a while back that Lessig has almost convinced him that people need a “right to remix”. But I didn’t think it would be fair to quote him on ideas in progress, and his categories are a solid application of freedom of use to cultural works.
    Remix is different from simple wholesale copying. If copyleft embodies the social contract of hacking at the AI Lab, then the social contract of contemporary remixing is Negativland’s “Transformative Fair Use”. Copyleft is a superset of this that includes cover versions, fan sharing, and other uses beyond remixing.
    The ISP tax is a form of ACS (Alternative Compensation System), as proposed by William Fisher in his book “Promises To Keep” which I recommend for a good description of the economics of the music industry (David Meme of Loca recommended it to me). I am not against such a system, I would gladly pay to get money to musicians I like each month as there are far too many bands that I seem to be one of the last ten fans on the planet for, but I’m not convinced that such a system is technically workable or will benefit musicians equitably.
    The second post in this series is going to be on the reputational side of allowing derivatives of cultural works, which was the other half of your original question, but that might not be for a week or so as I have a review and a book chapter to finish next week.

  3. “Oh wow. Thank you for your kind words. Now just make sure you get rich off t-shirt sales otherwise I will feel soooo guilty. ;-)”
    haha Well, I think I’ve figured out how CC BY-SA can make economic sense for me, not only as a songwriter but also as a performing musician. (Most of my music income in the past year has actually come from playing violin with other singer/songwriters, for example.) I’m still trying to do some research and reflection on economic models for other types of artists, but I have an idea of how I am going to start out.
    “Copyleft is a superset of this that includes cover versions, fan sharing, and other uses beyond remixing.”
    True, I’ve become more sensitive to that recently.
    I’ve been back and forth on the idea of an ISP tax as an alternative means of compensation, but Michael Geist and Mike Masnick have raised some additional doubts in my mind. For example, what if other industries started demanding the same thing, such as newspapers who feel that Google and Craigslist are impacting their revenue streams? It seems to be a bit of a dangerous precedent to set.
    Though, I do think that discussion about such a system is a step in the right direction. It draws people’s attention to trying to find ways to address a reality rather than pretending that online music sharing can somehow be stamped out. *shrugs*
    Thanks for the book suggestion!

  4. It seems that part 2 does not exist, though claims that contains some of it.
    Regarding NC (inclusive of NC-SA), I think you (and its critics generally) underestimate its effectiveness at preserving existing censorship-based revenue and the effectiveness of SA as a substitute for same. Yes, filesharing is alleged to be at fault for losses in recording industry revenue, but evidence for this allegation continues to be severely lacking; see for a recent survey. At the same time compulsory and business licensing, both of which are completely unaffected by filesharing, remain large revenue streams for industry. NC retains those — though some collecting societies still won’t have it, but that is changing over time.
    So NC removes impossible-to-enforce-except-through-police-state censorship of filesharing while maintaining the ability to extort from business. This seems like a step forward in terms of both free speech and efficiency, in the static case anyway. One can still argue that NC is suboptimal because it delays (which can mean prevents) nondiscriminatory free culture from winning.
    Now an ISP tax — there’s something that is overly complex and delays free speech winning. Jurisdictions are beginning to declare the Internet a basic human right — and we want to put a levy on access in order to make up for a revenue loss for which there is no evidence and no fundamental right of the recording industry to maintain? #FAIL
    Why do I think a levy is worse than NC (admitting that both less than desired)? Because NC is voluntary and more importantly possible to move away from incrementally. A levy will be politically determined and once in place, only get worse due to regulatory capture. Furthermore, NC can help move norms to what ought to be the case — complete legalization of noncommercial sharing, sans any control or tax. (And not precluding eventual complete legalization of all sharing, in the fullness of time.)