In their zeal to denounce certain license clauses, some free culture advocates have fallen back on definitions to advocate their cause. But such definitions make hasty assumptions about the nature of cultural practice. It's just not a proper movement without semantic squabbles. Where would Linux be without the snappy FLOSS (Free Libre Open Source Software) catchphrase? Anyone worried that Creative Commons have the whole free culture thing sewn up will be reassured to know that a healthy debate about creative and cultural freedom is finally starting to open up. So just what is free culture, and why should anyone care?

It's something I have touched on in these pages before, for example when analysing the BBC's Creative Archive License. By prohibiting any kind of use outside of the UK, and commercial use within the UK, many claim it is "non free". A movement hardly wants to be praising the very things it militates against.

Spurred on by these debates, in the past year there have been two prominent attempts to write a kind of Free Culture Definition to parallel that of the Free Software Foundation (FSF).

The first, the Free Content Definition, has been collaboratively written on a wiki, led by a few notables including Benjamin Mako Hill of Debian fame. It is broadly similar to the FSF's "four freedoms" definition, guaranteeing the freedom to study, apply, modify and redistribute content, whilst allowing restrictions that require attribution, the availability of transparent copies, a "share alike" clause and the absence of technical restrictions.

The second is the the Open Knowledge Definition, which more closely resembles Debian's Free Software Guidelines. It lays down 11 conditions that determine whether or not a work is "open".

Both definitions take a controversial stand on key issues such as the prevention of commercial use, which isn't deemed to be free or open. They directly challenge Creative Commons to reasess its approach, which is based upon providing licenses to fit the needs of creators rather than taking a clear stance on a particular set of freedoms by advocating, for example, the Attribution-ShareAlike (CC BY-SA) license.

The result of their approach is that most licensors are using the NonCommercial license, even though few are likely to ever make any money as a result of that licensing choice, and will hugely restrict others' cultural freedoms in doing so.

Creative Commons aims to balance the rights of creators and their financial needs, with the creative opportunities of the public. Any use of the word "freedom" is meant purely in an enabling sense, as in "I have more creative freedom", rather than in a rights sense, as in "I am free to copy this work". The latter sense is closer to that used in the Free Content Definition, and by the FSF.

Of course the idea that somebody could claim ownership of a concept such as "free", or for that matter "open", "libre" and any number of other candidates, is preposterous. A brief look at the history of political philosophy will quickly put these debates in context. But for the groups concerned, definitions are an important way of having an impact beyond the pseudo-academic circles in which they debate.

The dominance of the GPL, and the compatibility of many other popular licenses, guarantees that code can be effectively shared. This is all the more important in other cultural domains, where people may mix many more pieces of work than usually happens with code, and where fair dealing and fair use exceptions are often insufficient. This is only possible when people rally around a single, relatively tight definition, which serves as the basis for licenses. It's a cost of using legal tools to win political, social or technical battles.

The oft-cited problems of incompatible Creative Commons licenses, each creating their own version of freedom in an isolated, ghettoised cultural space, provide a good argument against their choice agenda.

If the account given in Levy's groundbreaking book "Hackers" is to be believed, then Stallman was making a précising definition. This is where one clarifies how a word should be used by describing its range of application. Stallman claims that one should only use the phrase "free software" in the context of his four freedoms. In doing so, the man who Levy playfully describes as the Last True Hacker was simply codifying the old hacker culture in a legal tool that would defend and extend it.

Free culture in action. 'Remix Broad Street' by Tom Chance is remixed into 'Canto (for Evie)' by Rob Myers, which is in turn mixed into '1020' By Timur Sahin. All released under the Creative Commons Attribution-ShareAlike 2.5 license.

There is no generally agreed understanding of cultural freedom amongst practitioners, though, and those writing definitions usually come at the subject from legal and technical backgrounds. Their definitions are stipulative, meaning that they are arbitrary rather than describing some real thing (such as a cat) or a common notion (such as the hackers' understanding of software freedom).

That, or they codify their notion of creative freedom, and then universalise their experience to claim it is a definitive notion of creative freedom.

The essence of the debate is a political struggle between competing stipulative definitions. Critics of Stallman could lay the same charge at the feet of the Free Software Foundation, arguing that Microsoft's definition of "free software" (usually meaning free of charge) is just as valid as Stallman's. The question is not who is right, but whose definition is more compelling.

But there is a major flaw with the free culture definitions. If we consider the GPL to be a legal tool that guarantees a hacker's right to sit at a computer and code, then we find the free culture definitions lacking an aim. They seem to be an attempt to simply port hacker culture and free software definitions to "content" or "knowledge", without any apparent reason for thinking that the concepts are analogous.

They don't start from an ethical claim, as Stallman does, nor from a claim about superior techniques, as with Open Source. Rather, they begin with the unstated assumption that copyleft is universally good.

What we need is a different kind of definition, one that stipulates general features of cultural freedom rather than conditions upon licenses (a bit like describing a hacker ethic before writing the GPL).

With such a definition, the free culture movement can work on tools such as licenses, business models and software that expand cultural freedom, and move closer to overcoming barriers such as excessively restrictive copyright laws and DRM technologies.

The definition would also help broaden and deepen discussion about the meaning of cultural freedom, and the applicability of copyleft to different cultural contexts. Maybe we should allow creators to restrict commercial use, just with better legal tools than those currently offered by Creative Commons? Maybe licenses are simply a distraction from more important issues?

This stipulative definition could do worse than to start with the FSF's four freedoms. They are broad enough to avoid the trap of thinking copyleft is universally applicable, but specific enough to open up debate. How can we enlarge people's freedom to consume and produce, to share and learn, or guarantee their right to do so? Crucially, why should we do this?

Through careful conceptual analysis and the use of examples we can begin to unravel these four core concepts. For example, it is right and proper that we consider the relatively passive activity of consumption as being valuable, so long as we remember that people are also creative. There's nothing wrong with wanting to sit down and watch a film, but the experience is richer if we discuss the film with friends and family afterwards.

We begin to realise our potential, and develop more as citizens and human beings, if we can then learn the techniques of the film makers, produce our own mashups or parodies, and share our skills and works with others. The four concepts encompass various basic dimensions of cultural activity without making hasty claims about specific issues such as commercial use.

This kind of definition also allows room for important debates without implying that certain people are outside of the free culture movement, such as those who reject copyright altogether and those who use non-commercial licenses. Proprietary software companies may try to argue that freeware falls inside such a permissive definition, but then oil companies pretend to be green on the basis of a tiny investment in fuel cells. If our case is sufficiently compelling, if we can genuinely enlarge creative freedom for enough people, we needn't lose sleep over attempts by Time Warner to hijack the terminology.

Creative Commons may be creating all sorts of problems with their incompatible licenses. Their desperate desire to be seen as promoting a commercially viable model leads them to endorse some pretty dubious practices, and to largely ignore huge areas of interesting cultural activity. But their choice agenda conceals an important truth: we don't understand cultural practices enough to say "this is free culture, whereas this is not". We should see their numerous licenses as an odious complexity, on the one hand, but also a useful experiment on the other.

What I propose is really quite simple: let's explore cultural activity further, experiment with different tools, and all the while debate the application of a "free culture definition" that starts from the four freedoms, to consume, produce, share and learn. We may rally around the CC BY-SA license for practical reasons, just as hackers rallied around the GPL, but we shouldn't mistake that for a definitive statement about ethics. The freedom to err in our understanding of culture is just as important as the ethic we construct.

This article was published in issue 68 of Linux User & Developer.