Free Culture

Copyleft And Marxism

I am a militant liberal, possibly a socialist libertarian. My political motto is “beware teleology”, so I regard pluralism as the least worst option for society given what happens when societies fall under ideological closures. Including Neoliberalism. Pluralism has come to be the basis of my ongoing interest in copyleft. I believe it strongly supports pluralism. I am nonetheless not delusional regarding the scope and socioeconomic effects of copyleft. It is not a panacea, and I think that Negativland’s Expanded Fair Use might be a better baseline for creative freedom in bourgeois capitalist society were it not for the practicalities.

My research over the last few years has only strengthened my regard for Stallmanism. This research has taken me through venture capitalism, the economics of Web 2.0, neoliberal ethics, the history of the enclosure and allotment movements, the levellers and diggers, 1968 in France, LIP self-management, various anarchist texts, the production of Classical Greek pottery, the history of collectives and LETS schemes, and all sorts of other historical and sociopolitical/economic moments. Oh and Lessig’s writing and that of Isaiah Berlin and of Mackenzie Wark. I have maintained my interest in copyleft not due to doctrinal monomania but to being unable to find anything better given current historical conditions.

I’m writing this to try to gather my thoughts prior to responding to someone who has objected to my criticism of their essay. They have accused me of regarding copyleft as perfect and immutable and believing that anyone who doesn’t share this opinion just needs to read the right books. I have taken several days to consider this and have come to the conclusion that my position will be practically indiscernible from that position for them, but that this is a mischaracterisation of what I believe.

I do believe that copyleft will support the ends of socialists and marxists within society better than a licensing scheme that seeks to refuse “the commons” to “capitalist exploitation of wage labour”. I believe this because of my long consideration of the self-defeating effects of the Non-Commercial Creative Commons Licence and other discriminatory licenses, and because I don’t believe that capital is stupid enough not to start selling picks and shovels to the workers who will be the only permitted users of such a licence if one is ever written.

I also believe that there is probably some interest for marxist students of the property question in essays titled “Why Software Should Not Have Owners” and “Did You Say Intellectual Property? It’s A Seductive Mirage”.

But I am interested to hear if any of my more marxisant readers think I am talking bollocks, and if so exactly why. Anyone?

Technorati Tags: ,

13 replies on “Copyleft And Marxism”

Copyright and patent constitute the unethical suspension of our liberty, and violation of our privacy.
Copyleft goes a long way toward neutralising copyright.
So, don’t mistake the instinctive recognition of the goodness of copyleft, as arising out of copyright enablement, or the precise terms of copyleft licences.
Copyleft is a mere tool, not an ethical principle.
Decide carefully if you believe so much in copyleft that you’d recreate copyright in order to enable it. Do you really need law that suspends people’s liberty to share & build upon other artist’s published works? Do you really need law that will even prevent people making copies or derivatives in the privacy of their own homes?
So, are you a copyleft fanatic or a copyright abolitionist?
A more subtle contention exists between those copyright abolitionists who differ as to whether they’d also abolish intellectual property.
IP Communists: Unlike material works, all intellectual works belong to the community and it is wrong for any individual to attempt to hoard works, prevent access to them, or contrive artificial scarcity (for commercial gain). E.g. “Books may belong to the purchaser, but the words belong to the world. Private libraries may retain their books, but they may not deny access to the public.”
IP Naturalists: Intellectual works, like material works, are the natural property of their authors or purchasers, who have a right to control access, or realise the value of those works in a free market. E.g. “When you buy a book, you own the words within it as much as the paper and can do what you want with them. And when you sell a book, you sell the words too.”.
Might as well mention the IP maximalists who’d consolidate copyright and patent into a more coherent (far more draconian) embodiment of the same principles (ideas exist as property independently of their realisation).
IP Maximalists: Ownership of an intellectual work may be retained (and transferred), even though the ownership of the physical medium in which it exists may be transferred. Ownership initially attaches to the producer of the original expression and is conferred by communication and similarity. No material object in which a similar intellectual work manifests may be produced without the original owner’s consent. Communication of similar intellectual works are likewise subject to consent. E.g. “If it looks like Mickey Mouse, or it was copied from Mickey Mouse, then it belongs to Disney, or his assigns/heirs”.
You’re evidently not an IP Maximalist, and are probably against copyright and patent (unless you see them as indispensible for copyleft), but are you also against the very notion of being able to treat intellectual works (realisations of human thought) as property?
I am not a Stallmanite. I do believe that software should have owners, and that intellectual property is a perfectly wholesome concept.
One can be against monopolies without being against property. I fear it is easy for people to mistake the natural monopoly of property as being just as iniquitous as artificial monopolies (of copyright and patent).
“I own the software I write and the software I’ve purchased, and there’s no way I’m going to let any damn ‘IP is theft!’ commie snatch my IP from me. If you covet my IP, offer to pay for it.”

Hey Rob, I read your blog and like the stuff you right. However, it’s stuff like this that keeps the right wingers (like me) away from copyleft, thinking it’s only for communists. I think that copyleft can create a far better capitalist society than we are used to, democracizing (is that a word?) the ip playground, and changing capitalism to what it’s suppossed to be about: fair competition. I’m not telling you to change what you post. I just thought I’d add a quick capitalist perspective.

Crosbie – Yes copyleft is just a tool. If there was something that was shown to be better for achieving the same ends I would use it instead. This is the hacker approach: you need running code.
I must confess I haven’t thought how far I would abolish copyright. Given that, for software, the public domain is not sufficient to create a public domain, I wouldn’t abolish all software copyrights unless software object code became as editable as source code. For culture I regard properly applied Fair Use as *maybe* sufficient (voluntary copyleft is better, but as a social baseline Negativland’s Expanded Fair Use might be acceptable).
As for opposing monopolies but supporting property, this is (close to) Lessig’s position.
There are many interesting positions *within* Free Software / Free Culture debate that touch on “the property question”, and that could could usefully be twinned or contrasted with and extended by positions within Marxist discourse if one so wishes.
Within pluralism marxists can pursue their end of revolution. If they attempt to turn copyleft into an instrument only of revolution this will fail. We have seen this time and time again with anti-military, anti-capitalist, anti-whatever licenses. To classify a class-discriminatory licence with good intentions in this way is not a failure of imagination or copyleft monomania. It is to learn from History, and to maintain the irony that makes copyleft so powerful.

The public domain is only insufficient today because copyright can appropriate from it.
Without copyright no-one can enclose public domain works by making a slight adjustment (minimum necessary to constitute originality in derivation) and slapping (c) on the result.
So, without copyright, if the source to s/w is published, then the published source to a derivative is just as free as the source it was based upon.
As to the idea that source code must be compulsorily disclosed check out my comments on Luis Villa’s blog here:
Source code is only kept secret for proprietary s/w because it assists the copyright based revenue model. Without copyright there is no incentive to keep the source secret after the s/w has been paid for.

Crosbie – I believe that copyright is required to make source disclosure enforceable by the GPL. Without this you could not have a requirement of disclosure. This would allow binary-only releases of derivatives which, for current compiled languages, would make further modification impractical. Is there another, copyright-free mechanism that could achieve this?

Conley – My point here is that I believe that copyleft very definitely *isn’t* only for communists, and that this is the benefit of copyleft to communists. Under pluralism, a marxist and a capitalist pursuing their own ends within society benefits society as a whole. And each other.
I’m just asking whether I’m correct to believe this as I’ve been challenged on it by a marxist critic. If a capitalist had challenged me then I’d be phrasing this in more markety terms. 😉

Rob, if there’s no copyright, why do you need to compel disclosure of source code?
Or putting it another way, if there was no copyright, why on earth would anyone pay for just a binary?
In a free market where a binary is free to copy and free to compile from the source code, why is anyone going to buy a binary?
Let’s remember, the source code is FAR more valuable than the binary. Therefore, the revenue available from sale of the source is going to dwarf the sale of the binary. And in both cases the same amount of work had to be peformed. The binary is simply an obfuscated version of the same source code.
So if you’re a vendor stuck trying to sell a binary that will quickly diminish in price from $10 to $0, or a vendor giving away a binary in order to advertise source code for $1,000, the latter vendor is going to outcompete against the former (even if the former hopes to sell their source eventually, for $100). This is because the former vendor is not using a free binary to promote sale of their work to a far larger audience for a larger amount of revenue.
If software users value source code, then the source code should have a price. There’s no reason it should be confiscated from its author without compensation simply because the author promoted their work via a free binary demo.
Copies are free of charge, and binaries are free of charge, but that doesn’t mean the author shouldn’t get paid for the publication of their source code. This is their labour and they should be able to exchange it for whatever they can get in a free market.
This is the difference between IP naturalism and IP communism. Both are against artificial monopolies, but the latter also believe an author’s work should be confiscated from them without compensation (simply because they had the audacity to publish a binary without source).
So, no copyleft doesn’t represent a meeting of capitalism and communism, it represents only a shared antipathy toward artificial, state granted monopolies.
IP naturalists still believe an author owns their intellectual works like any material property, whereas IP communists believe that an author’s works automatically become the property of the community for the benefit of all.
The key difference is on the compulsion of disclosure. IP naturalists consider this a violation of the human right to privacy, IP communists consider that non-disclosure represents theft from the community. However, we all agree that all artists should be free to share/buy/sell, study, copy, and build upon anything that is published, and thus support copyleft licenses that preserve the restoration of these liberties for all.
IP maximalists/mercenaries don’t give a damn about ethics, and simply want to exploit, protect and extend the privileges they’ve been granted, and become used to over the last few centuries (so much so they consider them ‘rights’).

Crosbie – The source is (economically) worthless to the average user, who is not a hacker. The source is far more valuable to a(n economic) competitor. Remember that in American law software could be a trade secret before it was copyrightable, and that the desire for copyrightable software was fulfilled with the 1976 US copyright act.
BSD licenced software ends up in proprietary software all the time. I believe that even without copyright the incentive for not giving improvements to economic rivals would mean that developers would be tempted not to disclose source code improvements that they regarded as giving them an economic advantage. I do not believe that proprietary software is a product of software copyright, quite the opposite. And we cannot simultaneously argue that people will pay for Free Software and that there is no value in software without copyright.
I agree absolutely that the compass for all this should be ethics, and I do find your principles of truth and privacy compelling.

Value is different from utility.
The user highly values the software, irrespective of whether they need a copy of the binary or the source for their purposes.
However, when it comes to the time for the author of the software to sell their software, the issue we’re considering is whether they’ll sell just a binary or the source code. The party offering money in exchange for the software is the one most interested in getting value for money. They will say “Demonstrate that you are able to deliver the software, and then we will exchange money for it”. Now, it is this purchaser that is aware of the difference in value between a binary and the source code. They are not going to pay anywhere near the same price for a binary vs the source.
Now consider the author. They have done X days work to produce the software and have potential competitors who may have written comparably valuable software. Are they going to sell a binary for $10, or the source for $1,000? Perhaps they’ll do the former – in case they need to do a bit of marketing to find the latter deal. But, you must ask yourself, “Why are they going to do a $1,000 worth of work, and be happy with $10 for it?”.
“They’ll keep on selling copies of the binary for $10 each” you say…
No they won’t, because copyright has been abolished. Anyone can copy the binary and give it away for nothing, it can be instantaneously diffused across the planet. They’ll be lucky to sell ten copies at $10.
“But, people have a right to the source!”
No they don’t. Some people will undoubtedly highly covet the source code, but that desire doesn’t constitute a right to confiscate it from the author.
Those that want the source need to go back to the author and say “Ok, thousands of people have been using free copies of the binary version of your software that you sold one day to someone for $10. We’d now like the source code to it because we’d like to modify it.”
Given competition they’ll also say: “Although, there are umpteen other authors who’ve produced comparable software, we have a hunch that your software is slightly better, we’d like to offer you $1,000 for the source, but bear in mind we could offer someone else $900 for something similar.”
And this is not extortion. It’s called trade in a free market.
The author has laboured to produce software, and is interested in exchanging the value of this labour for whatever price they can get in a free market.
Remember, they have to compete with others. They will be unlikely to sell their software for significantly more than the cost of the labour it would take anyone else to produce.
This is why binaries will tend to be free promotional copies for the source code and/or enhancement services of the original author.
It is also why it would not only be unethical to compel the disclosure of source code, but unnecessary, given a free market in which source code can be sold.
If you have a binary and there’s no source code for it, it should mean that it hasn’t been purchased/published yet. It should not mean that there’s an author who will shortly be prosecuted, fined/imprisoned, their computers seized/searched, and their source code compulsorily published.
Abolish copyright, but do not replace it with something just as unethical.

On your point about the BSD, I think you’ve accidentally fallen into the trap of mistaking its ‘abdication of copyright privilege’ as equivalent to ‘copyright abolition’.
The GPL is far closer to ‘copyright abolition’ than the BSD, and differs from copyright abolition only in the requirement that published derivatives not be obfuscated.
The BSD is entirely irrelevant to this discussion, it bears no resemblance to any issue at hand.
My concern is that because the GPL restores the liberty suspended by copyright (which is jolly good) and preserves its restoration (excellent), people think that the GPL is perfect and holy in every respect and should be enshrined in law.
I’m trying to point out that actually, the GPL should only repair copyright’s damage. It should not attempt to extend ‘liberty’ beyond the public domain into the private domain.
Instead of exploiting copyright’s ability to police the private domain, it should completely nullify it. Unfortunately, despite nullifying it to a large extent, the GPL does reserve a small exploitation by its compulsory disclosure of private source code in published derivatives. Now, this is consistent with using copyright against itself only if it is directed against those who would seek to re-suspend someone’s liberty or violate their private domain. However, people are beginning to believe that they have a right to seize the source code of anyone who publishes a GPL binary (mostly, the same people who believe copyright is a right rather than an unethical privilege).
This belief in a right to source code also betrays the motivation behind the Affero license and the AGPL.
Stallman’s four free freedoms should have been called three freedoms and a want.
Freedom 0: To run/use.
Want 1: The source code
Freedom 2: To copy/share.
Freedom 3: To modify.
Copyright doesn’t suspend freedom 0, although some licenses may exchange constraint upon use in exchange for partial restoration of freedom to copy/modify. Copyright only suspends freedoms 2 and 3.
The thing that suspends ‘want 1’ is not copyright, but privacy. The thing to remedy this is not compulsion or confiscation, but purchase (or gift).
This leads us to a somewhat more ethical set of principles concerning freedom:
Freedom 0: To use purchased/gifted software in any way.
Freedom 1: To be able to obtain source code for study or modification via a free market in which authors can offer to exchange their source code for money/goodwill offered by those who want it.
Freedom 2: To copy, share, sell, redistribute purchased/gifted software.
Freedom 3: To study, reverse engineer, or modify puchased/gifted binaries or source code, and use/share/sell/publish derivatives, etc.
All the above now need make no impairment to truth or privacy, nor should they.
No doubt one could also enumerate several things that arise out of rights to truth and privacy, but that’s for another day.
We can call truth and privacy respecting software that adheres to this revised set of ‘freedoms’ as ‘Ethical Software’.
‘Ethical Software’ as in ‘Software published in a free market by an author who may speak freely, is accurately attributed, and does not have their privacy violated except in order to protect life’.

Are you arguing for a kind of Street Performer Protocol for code in the absence of copyright? There are pieces of software and cultural works that I personally would gladly take out a second mortgage to get hold of the source code to (notably the sources for AARON, the sequencer and sampler files for Floodland, and the electronic source files of preparatory work for art by Julian Opie and Fiona Rae). In the absence of the ability to prevent propagation of built versions of works, I can see that this would provide an attractive potential source of revenue for producers.
In this scenario the author of the software will be reluctant to release their source code for fear of competitors adding value to it and not releasing the modifications. What Greg London refers to as a fork. This is why I compare the situation that would emerge in the absence of copyright to BSD rather than GPL; the GPL ensures that modifications are published, BSD encourages people to seek an economic advantage by not publishing. My concern is that fear of competition will override any calls from the unwashed for source code that they lack the skills to hack. Under SPP you can pre-charge for just binaries. By definition, anyone who appreciates the value of source code enough to ask for it will be at least a potential competitor, or will be at least potentially likely to seek a cheaper provider of modifications to it.
I agree that privacy is vitally important. The elimination of private life is a mark of totalitarianism. I forget whether the GPL’s respect for privacy is principled or pragmatic. Creative Commons licences do not acknowledge private use (presumably this would be part of Fair Use where that exists). My sketchbook should not become BY-SA because I collage in a BY-SA postcard (although if I publish it then it should become BY-SA, again modulo Fair Use where it exists). I would be interested to see a discussion of whether and how private use and publication should affect non-software copyleft.
Where I use proprietary software in a private capacity, I am using it in a public capacity relative to the originator of the software; I am part of their public. The binary executable derivative of proprietary source code is the public tip of a private iceberg. The privacy of the author of the software should not trump my privacy. Proprietary software makes my private use of software a public concern whether it works correctly or not. Proprietary software survives the abolition of copyright; it becomes any publicly released software with private source code.
The creator of a piece of software is legally and practically an author. But they are also described as an engineer, with their work regarded as engineering or even mathematics. It is difficult to construct a Romantic conception of software authors when two pieces of software can be functionally equivalent despite being written in different languages using different idioms with different interfaces. Software falls under copyright but in the US also falls under patents, and the GPL’s disclosure requirement is patent-like in effect. Software is an invention, not an expression.
Software is also inherently derivative, far more obviously so than literature and other texts. If we accept that linking is derivation then far from being original authors, software producers are producing palimpsests. Software is not inherently public but like science (rather than alchemy, diaries, or home movies) it is inherently the product of public resources.
If all programming was done in interpreted languages then the mere absence of copyright be ideal (modulo code obfuscation). If tools existed to modify binary code with the clarity of (good) source code, this would not be an issue. But binaries are still the norm for desktop applications, and decompiled code lacks comments and meaningful names.
The author (rather than engineer or artist) of a piece of software is the author of the source code. The source code shows the author’s hand and personality, it is the sweat of their brow as much as this is possible for an invention. A binary executable is a mechanical transformation of the actual text of the software. Whatever its practical advantages it is an obfuscation, a precis, a representation, a derivative of the actual work. If software engineers wish to be seen as authors they must publish their texts as authors do, not release the executable equivalent of uncopyable eight-track audiobook abridgements.
I do not believe that the GPL is holy and should be enshrined in law. I regard copyleft as the “least worst” solution to a technical problem for a principled position (the ideology of Stallmanism) wih great practical benefits for individuals within society. I do greatly appreciate copyleft but I recognize that some projects would be harmful and simply fail as copyleft projects. OCAL should be public domain for example, and although I do believe that OSM should be copyleft I find the arguments of those who argue for different approaches coherent if not persuasive.
The absence of source for cultural works is a loss rather than a problem. The absence of source for software is a problem. If software is a work of authorship then the product of that authorship is the text of the source code, and it is that which should be published. This should not be mandated by law, but I am uneasy about making it legally impossible to ensure this.

(I have only addressed your first paragraph, I’ll get to the rest in due course)
I am not arguing for any particular revenue mechanism in the absence of copyright, I am arguing that if people are not careful they will remove an entire class of revenue mechanisms because they make the mistake of thinking that all constraints upon IP are bad constraints.
So yes, revenue is possible in the absence of copyright, but in order to enjoy the same kind of commerce as non-digital art, it still requires the recognition of intellectual property (that intellectual work can be owned, bought and sold).
There are two problems resulting from the failure of imagination of a world without copyright, and the belief that revenue is impossible or insufficient.
1) Proprietary: Because revenue is impossible, we must therefore retain copyright.
2) Free: If revenue is impossible, we don’t need to worry about it, and can treat all constraints upon intellectual work as unnecessary constraints upon freedom.
It’s this latter camp that are IP communists. “All work belongs to the community, therefore anything that keeps it from us is wrong, i.e. copyright, patent, and… privacy.”
However, privacy is vital to enable authors and artists to sell their uncommissioned, but completed and demonstrable work. Without privacy one can only be commissioned in advance of the work. One is forced to give one’s initial portfolio of masterpieces away in order to attract commissions.
If we retain and protect privacy, naturally and ethically curtailing the public’s freedom, then artists can demonstrate their uncommissioned, as yet unsold work and offer it for sale in a free market (without interference from the monopolies of copyright and patent).
For example, but for the obligation to release source code, one could take a GPL work, produce a great improvement, give the binary away or provide remote access to it, and then sell the source code. It is better to think of this in a world without copyright, i.e. take a published work, produce a great improvement, give the binary away, and then sell the source. Remember, the author has done a lot of work and has not yet been paid. Why should their source code be confiscated from them and published without compensation?
It is like an oil painter giving away a photo of their oil painting in order to promote and sell the painting. The painting is their property, why should it be confiscated from them simply because the public covets the source for the photo? Moreover, why should it matter if the artist based their painting on someone else’s?
If you want to keep it digital, let’s say a musician gives away a low bit rate MP3 of a studio recording in order to promote the sale of the digital master. Is the master forfeit? And should it depend whether this recording comprises samples from other musicians’ published masters?
This is the new struggle bubbling under. IP naturalists vs IP communists.
Natural rights vs ‘IP is theft’.
Anti-Monopoly vs Anti-Property.

(Addressing a few more paragraphs of your comment)
If by ‘proprietary software survives the abolition of copyright’ you mean ownershp of intellectual property continues, then yes I agree. But, then this is natural.
If I sell you a photo of my house, what is wrong with me keeping the designs to my house private? Either you want to buy the photo or you want to buy the designs, or both. As long as what is offered in any exchange is understood by both parties, then all is well and good.
‘Proprietary software’ can also refer to the copyright based business model, that enables the sale of uncopyable binaries and thus obviates and disincentivises sale of source code.
As I have already been trying to argue, it is only copyright and not entitlement to own intellectual property, that keeps source code off the market. So, yes, in the presence of copyright, it may be necessary to oblige release of source via the GPL, to demonstrate the viability of a market for source code. However, in the absence of copyright, source code will reappear in the marketplace, simply because it represents what is being sold. The source is the labour. When you buy software you are buying the labour, therefore you will buy the source. With copyright, when you buy a binary you are paying well over the odds because of a monopoly. Without that monopoly, there’d be an unviable market for binaries.
If you want a house, you’ll pay for the house. If you just want to look at a house you may be happy with just a photo. It doesn’t matter what the relationship is between the photo and the house. Each of them can be bought and sold. One might say that the photo obfuscates the architect’s designs of the house’s exterior, but the motivation for supplying the photo instead of the house or its designs is irrelevant unless intended to deceive.
A binary can be regarded as a convenient means by which an author may demonstrate the work they have done. The demo may be valuable or useful in itself, but the fact that it isn’t as valuable as what it is promoting or demonstrating is not a deception, not an infringement of the audience’s liberty, nor an unwarranted denial of what is being demonstrated and offered for sale.
However, if I write a program with full source comments and inline documentation, and sell the source code to someone as ‘source code’, then it would be deceitful indeed had I obfuscated the source (stripped all the comments, randomised variable names, removed formatting). Then again, if I had explicitly stated that I was supplying software for demonstration in advance of sale and had thus obfuscated the source (because it was an interpreted language say), then this is like a photographer overlaying a ‘preview’ mark to their work.
Incidentally, I use author and artist interchangeably as ‘intellectual worker’ or ‘one who has, through the intellectual labour of thought rather than mere recollection, produced a work that may be apprended by the intellect of others’. So it doesn’t matter how similar their work is to anyone else’s whether by coincidence, inevitability, or similarity of experience or source materials. It is copyright that has created/embodied an obsession with originality. And yet we know today, that given enough monkeys there is very little new under the sun. Software engineers no longer consider ‘reinventing the wheel’ as something to be avoided and a failure when it happens, but as natural as convergent evolution.
So, frankly, I think software engineers are the last to pretend that originality is their primary aspiration and the criteria by which their work should be judged. If this disbars them from consideration as authors, so be it, but it no less diminishes their belief that they are the one’s who laboured to produce their work.
Of course, buyers and sellers of work should be accurate in their descriptions, e.g. how much work is actually being sold, and how much of it is available elsewhere, etc. We’re back to deception again. One can take someone else’s work make a minor alteration and pretend to far more work than was actually performed. I am not supporting plagiarism or misrepresentation, but prohibiting sale of binaries without source is an unethical sledgehammer even if it is a means to prevent deceit.
I would actually agree that the absence of source for cultural works is a problem. However, I believe that absence is caused by the discincentivising consequence of copyright enabling monopolised sale of inferior derivatives. Compulsion of disclosure is only possibly necessary in the presence of copyright, and I don’t actually believe it is necessary in copyright nullifying licenses either. However, if copyright and patent are abolished, compulsory disclosure of source is not only unnecessary for software or any class of intellectual work, but it would be just as invidious and unethical as copyright.
The best means of encouraging the publication of source materials is a free market in which those interested in the source materials can offer money in exchange. And a free market is one necessarily free of monopoly.

Comments are closed.