Peer production and Marxian communism: Contours of a new emerging mode of production

Introduction

The first section of this article discusses the emergence of PP’s forms of property, modes of cooperation and distribution, highlighting the historical novelty and significance of these forms. The second section analysis PP, comparing it to Marx’s definition of advanced communism. The third section aims at showing why PP is a new historical, though still emerging, mode of production, and not merely an epiphenomenon on the margins of capitalism. The fourth section deals with PP in the realm of material production. The fifth argues that PP can subvert capitalism only through a social revolution that transforms strategic means of production into commons. The article concludes with a few remarks on the contingency of such a revolution.

This article picks up many of the themes of the special issue of Capital & Class (issue 97, 2009, edited by Phoebe Moore and Athina Karatzogianni) on PP. While all contributors to the issue recognise the novelty of peer production, they differ in their evaluations of its relation to capitalism and its transformative capacities. Bauwens (2009) argues that although there is a symbiosis between PP and capitalism, the former may replace the latter. Orsi (2009) argues that PP (a gift economy), a state economy (public goods economy) and a market economy can be combined in a way that is beneficial to the majority of people. Dafermos and Söderberg (2009) argue that hacking is a form of class struggle that de-alienates labour.

Böhm and Land (2009) argue that creative-artistic labour may serve the interests of capital by reshaping the social in the image of capital. Moore and Taylor (2009), recognising the transformative capacity of PP, suggest that the volunteer work carried out in PP may serve the interest of post-Fordist capitalism. Karatzogianni and Michaelides (2009) highlight the roles of conflict and hierarchy in PP. Vadén and Suoranta (2009) argue that digital PP is currently circumscribed by capitalistic material production, and argue for a socialist transformation of society.

While acknowledging these contributions, this article has four distinctive features. First, it situates PP more explicitly in the framework of Marx’s theory of modes of production and their revolutionary transformations, on the one hand, and his understanding of advanced communism, on the other. Second, while recognising the current links and symbiosis between capitalism and PP, it highlights the deeper contradiction between them. Third, it highlights the role of 3D printing for the development of material PP. Fourth, it argues that PP can overthrow capitalism only if the strategic means of production (land, major sources of energy and raw material, and major technical infrastructures) are transformed into commons.

Two groundbreaking revolutions: The origins of PP

Since the history of contemporary PP has already been covered by others (Weber 2004: 20-53, 94-127; Söderberg 2008: 15-26; Raymond 2001; Stallman 2002), here, I limit myself only to two defining moments of this history: namely, the invention of GPL (general public license) and free software (FS) by Richard Stallman in 1984, and the invention of the system of online voluntary cooperation by Linus Torvalds in 1991. I contend that these two inventions are major revolutions, and that they comprise the two main features of PP. Stallman invented the particular PP form of property rights, and Torvalds invented the PP mode of cooperation.

Let us start with Stallman’s invention. Source code in software consists of human-readable algorithms written in a programming language such as Python, Java, C, etc. Source code is translated into machine code (binaries), which computers can read and execute. It is extremely difficult, and indeed next to impossible, to modify, customise or add new features to a program without having the source code. Source code cannot be extracted from binaries, and therefore access to it is necessary in order to make any change in a program (Buckman and Gay 2002: 3-5). In commercial software such as Microsoft’s, the source code is a commercial secret. In contrast, free software (FS) and open source software (OSS) release the source code, with full documentation on how it has been written. Richard Stallman (2002) relates the story of the invention of FS as follows: up until the 1970s, the small community of software developers liberally shared codes with one another. As a result, anyone could modify, customise and develop further programs written by others. The commercialisation of programming interrupted this practice. Capitalists imposed copyright restrictions, and obliged engineers to sign non-disclosure contracts. This provoked resistance among programmers, which was initiated and led by Richard Stallman. Stallman resigned from his job at MIT in 1984 in order to develop ‘free software’. In Stallman’s conception, ‘free’ does not necessarily mean that one cannot sell the software, but refers to the concept of freedom as liberty. In order to create the ‘Free Software Foundation’, Stallman invented the general public license (GPL), and wrote a very sophisticated code (GNU) that was released under GPL. GPL guaranteed four freedoms: the freedom to run the program for any purpose; the freedom to study and customise the program; freedom to redistribute copies either for free or for a reasonable price; and the freedom to change and improve the program (Stallman 2002: 20). Copyright law gives authors the right to specify the terms of use of their work. Stallman, using this right, included the so called ‘copyleft’ clause in the GPL in order to keep the derivatives of the code also free. According to this clause, any code that would include a component derived from a GPL licensed code must also be released entirely under a GPL license (Stallman 2002: 22-23). Copyleft can thus be understood as a dialectical negation of copyright, because it simultaneously, preserves and abolishes copyright. Therefore, GPL can only be enforced in countries in which copyright is enforceable.

The historical significance of GPL can be understood in two ways. First, it formulated for the first time in history a globally all-inclusive property right. Many people on Earth are currently unable to use this right because of a lack of access to computers and IT skills. But this is not a juridical matter, and by no means undermines the historical significance of Stallman’s invention. Stallman’s invention gives an important urgency to the demand that computers and IT education should be universally available for everyone.

Commons have existed since the inception of humanity, in various forms and among various civilizations (Marx 1965; Polanyi 1992; Ostrom 1990). But they have always been territorialised by belonging to particular communities, tribes and states. Hence, as a rule, outsiders were excluded (Marx 1965). Although seas/oceans and the atmosphere have also been a global commons, GPL created for the first time in history a juridical concept of a globally protected, deterritorialised humanly produced commons.

In the West, before the Renaissance, shamans, magician, priests and artisans tried to keep their expert knowledge secret, or to transfer it only to select individuals. The Renaissance broke with this tradition by making science a universal commons (David 2005), the production of which was governed by a communistic ethos (Merton 1979, 1996). Capitalist regimes of copyrights and patents contravened the Renaissance’s approach by fencing in particular forms of applied sciences (Boyle 1996). Stallman’s invention radically questioned these regimes by instigating a revolution in the realm of property rights.

Second, this revolution was staged in an era in which knowledge was becoming a paradigmatic force of production, and a major form of wealth in emerging informational capitalism (Castells 2010 [1996]; Benkler 2006). Hence, the ownership of knowledge has become a globally important bone of contention. Capitalists have always fenced in knowledge in order to extract rent. The dramatic expansion of increasingly draconian copyright restrictions since the 1980s reveals this tendency (Lessig 2005). In contrast, the majority of users and a considerable part of knowledge workers want knowledge to be a global commons. This is also in the interest of humanity at large. The GPL/GNU pioneered a legal-technical strategy for producing a global commons of knowledge and protecting it against capitalist invasion. In this sense, Stallman’s initiative marked a major turn in the social struggle over knowledge (Stallman 2002; Söderberg 2008). Stallman’s strategy was not merely negative. Rather, it aimed at producing and keeping knowledge as a collective property of humanity. Torvalds took the next ground-breaking step in 1991 by inventing a horizontal, network-based form of cooperation. He put the ‘Linux kernel’, a crude and incomplete base code, on the net, and invited anyone who was interested to participate in a collective effort to develop it further. (A kernel is the core of an operating system of a computer, which links the applications to the processing of data in the hardware.)

Copyleft allows commercial uses. A capitalist who uses knowledge released under GPL as a means of production will be able to produce at a lower cost than can his competitors. This enables him to make extra profit (surplus value) by selling the product more cheaply than his competitors, but above its production price, and by simultaneously taking a larger share of the market (Dafermos and Södeberg 2009). Therefore, Kleiner (2007) proposed the ‘copyfarleft’ licence, in which cooperatives of creative workers can use each other’s products for free, but privately owned companies must pay for the use of these products. The problem with this solution is that it leads to the exploitation of non-knowledge workers by knowledge workers. The intellectual products, as Kleiner observes, have no exchange value, because they can be copied at zero cost. Therefore, their prices are rents. The origin of this rent is the total surplus-value that is produced by the total wage-labour outside the knowledge economy. In Kleiner’s model, the capitalists who use PP products must transfer part or the whole of their extra profit to cooperatives of creative workers in the form of rent. But this makes cooperatives the exploiters of the total wage-labour, which produces surplus-value. The advantage of copyleft over copyfarleft is that it does not further the existing gaps between intellectual and manual workers. Furthermore, the free use of peer-produced knowledge by certain capitalists does not transfer exchange-value (surplus-value) from PP to the capitalist sector, because PP does not produce exchange-value at all. It only influences the distribution of the total surplus-value, produced outside PP, among capitalists. The major task of different sectors of the working class is to unite as a class, and not to bother about the competition among capitalists. In this sense, copyleft is more radical than copyfarleft.

Torvalds used Stallman’s GNU in writing the Linux kernel, and released it under the GPL licence. It didn’t take long before other developers began joining Torvalds, and ultimately, the cooperation of thousands of volunteers has made Linux a major program that has surpassed Microsoft in many areas (Raymond 2001; Weber 2004). Cooperation works as follows. There is no central division of tasks: each contributor chooses to work on problems of her own interest, and solutions are published and discussed on the community’s mailing list. Torvalds and his lieutenants coordinate the cooperation, and have the authority to choose between alternative paths of development. However, they try to accommodate the common will of the community, which is shaped through democratic discussions on the mailing list. Nevertheless, if an individual or a group disagrees, she/they can take the entire code and develop it in the direction she/they wish(es). This, in hackers’ terms, is called ‘forking’ (Raymond 2001; Weber 2004).

The historical significance of Torvalds’ invention can hardly be exaggerated. First, it is perhaps the most important revolution in the realm of the organisation of cooperation in human history. Second, as with Stallman’s invention, it bore the mark of its time. Such cooperation could only be organised using the internet: without the internet, it would be impossible.

The existing hacker movement and its cooperative culture, combined with invention of the internet, constituted favourable conditions for the emergence of Linux. Therefore, Torvalds’ role as the initiator of this mode of cooperation, though very significant, should not be exaggerated. Instead, it should be seen as a key moment in the hacker movement.

Production has always been mediated by cooperation among individuals. In pre-capitalist societies, cooperation took place on a small scale, except when despotic states coerced large numbers of people to build buildings and monuments, dig canals or work in mines. However, large-scale cooperation of this type was not based on a complex and sophisticated technical division of labour. It was merely the co-presence of a large number of workers who performed either identical tasks, or tasks that were more or less independent from each other. It was capitalism that created for first time a large-scale, complex and meticulous technical division of labour in which the work of each individual worker was a moment in the total integrated work process of a factory (Smith 1993; Marx 1976). What Torvalds invented was cooperation on a global scale, which transcends the capitalist division of labour. As we will see, however, this cooperation is based on the distribution of labour, rather than on the division of labour.

Stallman invented a model for the free global and de-territorialised distribution of software, and, indeed, of knowledge in general. However, the production of software remained local. A single engineer, or a few who cooperated in particular labs like that of UC Berkeley, would develop a new code more or less fully before releasing it. Then someone else worked on a new feature, developing it more or less fully, before releasing the more complex code. Although each successive product was the result of the combined efforts of all who had contributed to it, the cooperation among the contributors took a linear form.

Before the invention of the internet, engineers exchanged codes by exchanging floppy disks. Hence, the individuals’ (or small groups’) contributions to a given total work were disconnected from each other temporally and spatially. There was neither a division nor a distribution of labour among the contributors. The cooperation was the unintended consequence of the movement of software along time and space axes. Each new contributor started with the cumulative work of previous engineers, and added something new. As Marx (1981) noted, this was true of scientific cooperation in general. Linux, through the internet, revolutionised the organisation of cooperation. Now a basic incomplete, though working, kernel was released on the net, and different people could choose to work simultaneously on solving different problems in order to complete and expand the code. The organisation of cooperation took the form of a fully de-territorialised and gradually global network. Torvalds’ invention made conscious and coordinated simultaneous cooperation among a large number of producers scattered around the globe possible. This occurred through a distribution of labour which is radically different from the familiar division of labour on the factory shop floor (Weber 2004). First, there is no top-down centrally planned and managed division of labour: each person is free to choose to work on the tasks that most appeal to her. She is also free to work in times and places of her own choice, and at her own speed. In principle, anyone on the planet can choose to cooperate as long as she has access to a computer and the internet, and has the required competence. The result is the full temporal and spatial de-territorialisation of cooperation which abolishes the panoptical organisation of the labour process (Marx 1976; Braverman 1999; Foucault 1995). There is one form of centralised control, in that the development of the code is ultimately controlled by individual(s) who put the initial base code on the net. It could happen that different contributors might open alternative routes for the further development of the code. Initiator(s) of the code have the final authority to choose the direction of development, and this is usually done through extensive discussions. However, if someone is not happy with decisions of the initiators, she has the right to take the entire code and try to develop it in the direction she pleases (Raymond 2002; Weber 2004).

The PP mode of production and Marx’s advanced communism

In this section, I first summarise the main characteristics of PP, then show how PP is similar to what Marx described as advanced communism.

The combination of GPL with Torvalds’ mode of cooperation in Linux constitutes the core of a new PP mode of production (Benkler, 2006). This mode of production has the following features:

O’Neil (2009) shows that hierarchy plays an important role in the governance of PP (see also Weber 2004; Karatzogianni and Michaelides 2009). But it is an open question as to whether hierarchy is an inherent feature of the PP mode of organisation, or whether it is rather a result of the influence of the capitalist culture of power on PP.

Moglan (2003), Barbrook (2007) and Kliener (2010), among others, have related the internet and PP to the emergence of cyber communism. Marx’s description of ‘the advanced phase of communism’ anticipated PP in three ways (1978a, c).

The first is the absence of the logic of equivalence in both. There is no equivalence between the quantity of what someone puts into social production and what she takes from the total social product produced. In advanced communism, each individual contributes according to her abilities and uses the products according to her needs. This principle, though in a modified form, is already realised in PP digital production, since the contribution and amount of time spent on production are voluntary, and the use of the product is not conditioned on contribution. The nature of reciprocity in PP production is a novel phenomenon, and its full discussion requires a separate article. Here, I just point to a common mistake among scholars who have analysed PP. They have, wrongly, considered this reciprocity as a form of gift exchange (Raymond 2001; Benkler 2006; Barbrook 2007). Gift exchange is different from the reciprocity that takes place in PP in at least three respects:

The second is that PP, like Marx’s communism, negates alienation by transcending the division of labour and replacing labour with joyful and creative productive activity (Himanen 2001; Söderberg 2008; Marx 1978a: 529-53). I deal with this point below.

The third is that, like Marx’s advanced communism, PP’s de-territorialised and de-centred distribution of labour abolishes the top-down logic of organisation which is the logic of the management of capitalist enterprises, and that of the state. Hence, a fully fledged PP will abolish the state altogether.

The concept of PP is different from Hardt and Negri’s concept of ‘the common’. Hardt and Negri (2009) claim that the common is ubiquitously present everywhere, particularly in the realms of social interaction and production of knowledge, language, codes, information and affects (2009). PP, on the other hand, is emerging as islands within the capitalist social formation. Its generalisation to all branches of production will require a social revolution.

Terranova (2000) claims that PP is an integral part of the current capitalism because, she argues, the volunteer labour that contributes to it is part of the global immaterial labour that produces surplus value. This is a mistaken position. The cognitive labour does not, as shown above in the critique of the ‘copyfarleft’, produce exchange value and surplus value. Revenues of knowledge capitalism are rents extracted from the surplus value, which is produced outside cognitive capitalism. Cognitive workers produce knowledge which is immediately universal commons. Cognitive capitalism extracts rent from non-knowledge workers by fencing in these universal commons. This fencing can be characterised as a form of exploitation, but is not immediately a process of extraction of surplus value. PP undermines cognitive capitalism in two ways. First, it emancipates the cooperation from the yoke of capital. Second, by un-fencing the commons of knowledge, it also undermines the extraction of rent from non-cognitive workers. But this is not to deny that under current conditions, some capitalists use PP as a rent-extracting mechanism (see the discussion on ‘copyfarleft’, above).

This analysis begs the following questions:

PP as a historically new mode of production

Currently, PP only consists of small islands in the sea of capitalism. It is dominated by and dependent on capitalism in the following ways. First, a section of capitalists uses the commons of knowledge produced through PP to extract rent (Rigi, forthcoming). Second, most workers who contribute to PP for free live on wage labour. Third, the contributions of such workers to PP often involve customising PP products for the specific needs of the capitalist enterprises for which they work (Weber 2004). Fourth, major capitalist enterprises such as IBM take part in creating open-source software primarily because the open-source method results in much better software, and it would cost them far more if they had produced the same software separately. Fifth, there are companies, such as Red Hat, that contribute to open source software but make a profit from selling the additional accessory services to users. Sixth, material PP in the currently predominant capitalist economy is dependent on the commodity form for purchasing some parts of the material input and selling parts of its products.

However, despite these concrete links, there is on a deeper level of abstraction a fundamental contradiction between capitalism and PP. As argued above, PP has a totally different logic from that of capitalism, and has emerged in response to imperatives of IT productive forces which profoundly contradict the capitalist mode of production. PP also contains a collective and productive non-capitalist (and to some extent anti-capitalist) subjectivity (Davidson 2010).

Let us briefly sketch Marx’s conception of ‘mode of production’ and its major structural contradictions. Production is a process in which humans produce goods and services. The forces of production consist of humans, their knowledge/skills, instruments they use, the material they act upon, and other material conditions of production, such as energy, land, buildings, etc. Relations of production are ‘definite’ and ‘indispensable’ relations among humans, which correspond to the material stage of productive forces. Property relations are legal expressions of the relations of production (Marx 1978b). The mode of production is the totality of the forces of production and relations of production. Of central significance to Marx’s theory is the dual relation of correspondence/contradiction between forces of production and relations of production. The forces of production constitute the content, and the relations the form of any given mode of production. The form supports the growth of content, but at a particular stage of this growth it becomes a fetter to further growth, and therefore it must be overthrown through a social revolution.

I propose that Marx’s thesis can be applied to contemporary capitalism as follows: (1) Information technology (IT) and its corresponding form of general intellect (see below) are the paradigmatic productive forces of our time. (2) These forces have, historically, made capitalist relations of production obsolete. (3) PP is the social form that corresponds to IT and its general intellect. (4) Therefore, a social revolution for the transformation of capitalism to PP is the main historical agenda of our times. I warn the reader against an economistic, deterministic and evolutionist interpretation of this thesis. As I will argue in the last section of this article, there is nothing inevitable about this transformation. On the contrary: everything is contingent upon social struggle.

Knowledge and the internet are the central components of IT, and operate through the integration of various technologies such as micro-electronics, telecommunications, optoelectronics and computers and microbiology in a larger system (Castells 2010 [1996]: 70-72). Knowledge workers and their subjectivities are an important component of IT productive forces. The cognitive component of the total productive forces of a given society is its general intellect (Marx 1973; Virno 2007). General intellect is the hegemonic factor of IT productive forces, and therefore, the global hegemony of IT in production also means the global hegemony of general intellect (Hardt and Negri 2000; Virno; Castells 2010). The hegemony of general intellect does not mean that material production is insignificant, but that it is taking place, and will take place, increasingly, on the basis of the general intellect.

The information/knowledge and network structure of the net, which are the core aspects of IT (Castells 2010), inherently contradict the commodity form and the capitalist organisation of production. I will examine this below.

The contradiction between IT and the commodity form

Network logic requires that knowledge produced in each node of a globally integrated network should flow freely and horizontally in all directions to all other nodes. Knowledge is a non-rival good. It can be reproduced (copied) at no extra cost. It is also a universal good. The same piece of knowledge can, simultaneously, be used by everyone on this planet. Moreover, its use value is never depleted. Digital copying and the net have made the global transfer of knowledge in all directions at the speed of electronic signals a possibility. However, the capitalist mode of production prevents the free flow of knowledge in all directions. It is true that the capitalist mode of production, adapting itself to IT, has become global, and has increasingly adopted a network form (Castells 2010). However, the sum of all potential links in the net dramatically exceeds the sum of links in the global networks of capital (Hardt and Negri 2009). Hence, the potentials of knowledge and the net, which are the paradigmatic productive forces of our time, exceed the capitalist mode of production. Capital carves for itself a selected sub-network of the total net: the global network of accumulation of capital. The flow of capitalised knowledge is fenced within this selected subnet. Even within this subnet, the flow of knowledge is not free. This is because in the competition among different multinationals, significant forms of knowledge have increasingly become industrial secrets, and are jealously guarded by a small number of designers and engineers of particular enterprises. Furthermore, commoditised knowledge can only move from one hand to another (from one node to another node) if exchanged for money. In other words, the commodity form itself is a form of fencing. It is true that the investments of big corporations in research and development contribute to new discoveries and innovations. However, the commodity form prevents the growth of the general intellect on both global and national levels. The general intellect grows through the two following mutually supportive and interconnected uses of knowledge: innovations by talented workers, and the use and improvement of knowledge by the general public. The commodity form hampers the growth of general intellect by preventing many talented workers and the public around the world from accessing and using particular items of knowledge. Hence, from the point of view of the growth of general intellect, the commodity form should be replaced with the commons. This is also a requirement of justice and equality, as the removing of patents and copyrights will be an important step towards reducing social and global inequalities.

On a more fundamental level, knowledge is ontologically incompatible with the commodity form in the era of digital reproduction and the internet. It cannot be physically withheld, since it can be copied digitally and distributed for free on the net. Only the state’s coercive power can fence it in by enforcing copyrights and patents, or imposing other regulations. Therefore, the very ontology of knowledge has grown into an antagonistic contradiction with the commodity form. Its free movement is in the interests of the majority of humanity, and is in fact a requirement for the optimal growth of productive forces. This contradiction implies that the commodity form should be abolished. Indeed, GPL abolishes the commodity form, and makes knowledge a universal commons.

The contradiction between IT and the capitalist organisation of production

IT also profoundly contradicts the capitalist organisation of production on both macro and micro levels. On the macro level, it contradicts the macro division of labour by contradicting the commodity form. In capitalism, workers of different branches of production exchange their labour for wages, and the products of their labour become commodities owned by capitalists. It is only through the sale of commodities that the private labours of workers of various branches become social, and are combined together as the total social labour of society (Smith 1993; Marx 1976: 163-177). The net is an open-ended network in which every node (here, an immediate producer) can immediately and horizontally connect to any other node (here, an immediate producer). Therefore, knowledge that is published on the net is immediately and globally social. It is true that the signals that carry messages are produced through physical infrastructures, which are located in particular places. But this is external to the logic of the net itself. This is not to claim that flows on the net cannot be disrupted or controlled by those (state and capital) who control this infrastructure. They can simply switch off communication channels, monitor or filter them. Therefore, one should not underestimate the joint capacity of state/capital for controlling the net.

The structure of the net also implies that units of knowledge production can become de-territorialised, global, open-ended, horizontal and cooperative networks of direct producers. Linux and Wikipedia are good examples of such units. This makes the micro technical divisions of labour, and with them, the factory regime, superfluous. In this regime, capital designs the line of production and places different workers in different positions. Hence cooperation among workers is a product of capital (Marx 1976: 455-491). The inventions of modern machines allowed for the perfection of the technical division of labour (Smith 1993; Marx 1997). Taylorism, using scientific methods, established a full despotism over labour (Braverman 1999). The capitalist use of IT has increased the despotic micro-management of labour by capital (Tomaney 1994; Huws 2003). Labour is compartmentalised in different closed spaces, and is micro-managed by the representatives of capital. Managers control labour processes, allocate the tasks, and integrate the work of separate workers into a total work process. The worker is not free to choose the tasks she performs, nor to work at her own pace and in the times and places of her own choice. Since capital is not a necessary precondition for cooperation among cognitive workers, it resorts to despotic methods for subjugating them (Gorz 1990; Hardt and Negri 2000). This externalisation of despotism in turn intensifies the antagonism between labour and capital. Hence, from the point of view of productive forces, capital has become a superfluous external shell that should be overthrown. PP overthrows the despotism of capital and replaces it with voluntary cooperation among workers. This cooperation is globally de-territorialised in terms of both space and time.

The top-down technical division of labour fetters the growth of production of knowledge within the IT paradigm in two ways. First, it hampers efficiency. Second, it alienates workers by holding back their creativity and freedom. PP, by contrast, enhances efficiency by giving producers a space for materialising their passions and creativity. Brook (1975) shows that under centralised organisation, an increase in the number of engineers working on solving a particular problem, beyond a certain limit, will decrease efficiency. The reason, Brook argues, is that this creates unnecessary complexities of an exponential rate, due to problems of communication. Raymond demonstrated that this is not true of PP cooperation (Raymond 2001: 34-63). On the contrary, in this form of organisation the increase in the number of workers who work on a code increases the efficiency of production, and results in a much better product. The participation of a larger number of contributors is rather an advantage in the production of software. Any code inevitably produces errors (bugs) that can only be detected when the software is used. Therefore, producers of commercial software such as Microsoft conduct so-called Beta testing, before putting the software on the market. They give away early versions of software for free, to a limited number of users, in order to find the bugs. The problem with this method is that not all bugs are discovered, since the number of testers is limited. In open source, since the code is distributed for free, it is tested by a far greater number of users. As a result, more bugs are more quickly identified, reported and fixed collectively. In Raymond’s terms, ‘with enough eyeballs all bugs are shallow’. Moreover, PP cooperation attracts far more minds, some very talented, to work on different features. In the end, their collective product is far more sophisticated and efficient than the product of a smaller group of engineers who work for commercial companies.

The beauty of the PP mode of cooperation lies in the fact that this inherent efficiency is not only compatible with but results from the materialisation of the productive passions, freedom and creativity of producers. The contrast with the capitalist technical division of labour, where efficiency creates alienation, could not be more pronounced. The IT workers whom I interviewed in Berlin at the Chaos Computer Club’s camp in the summer of 2011 confirmed this view. They said that micro management of their work by the managers of commercial enterprises prevented free experimentation and innovation. They had to perform predetermined tasks assigned to them by managers instead of what they had a passion for. Accordingly, this made their jobs boring and hampered innovation. Pekka Himanen (2001) describes the hacker ethic as being characterised by passion, creativity, love for freedom and commitment to community. In the Linux mode of cooperation, requirements of efficiency converge naturally with such an ethic. Let us examine this closer. As mentioned above, a kernel is an embryonic working code that needs to be developed further. It can be completed in many different, and often unexpected ways. The system is always evolving through the experimentations and improvisations of the volunteers, who add new features according their own preferences. The unexpected and emergent nature of new features and the ways this modifies the system give a playful, though serious, adventurous and exploratory nature to the productive activity of each individual. The commitment to the social is also an organic requirement of production, and not artificially added to it, because the absolute precondition for this mode of cooperation is that the code always remains a universal commons. This model can be generalised to all knowledge production. A person who works on a particular project or a design can put the work in progress online and invite others to cooperate in completing it. In this system, there is a distribution of labour, because, while the final product is a result of the total work of the community of contributors, participating individuals choose to work on different supplementary features of a common project. Those capitalist enterprises that make codes commercial secrets cannot participate in this mode of cooperation. Instead, a few engineers design the total product, and managers plan and manage its production by assigning different tasks to different teams. Ready-made full design and central planning foreclose the possibility of other potentialities emerging.

To sum up, the capitalist form of ownership, organisation of production and distribution hampers the growth of IT productive forces in three ways: 1) it undermines efficiency; 2) it hampers the creativity of knowledge workers by restricting their freedom; and 3) the commodity form prevents knowledge from becoming a universal commons, and thereby hampers the growth of general intellect.

PP, by contrast, operates in the opposite direction, not only fulfilling the requirements of the growth of productive forces but also meeting quests for freedom, democracy and social justice. Hence, we can conclude that PP is an emerging historical mode of production and not an epiphenomenon on the fringe of capitalism.

With this point, we can now raise the question of whether and how PP is applicable to material production.

Can material production and distribution be organised through the PP mode of production?

A methodological note is necessary here. Since material P2P production is still in its initial phase, most of the arguments developed in this section are speculative and tentative hypotheses.

PP has been extended beyond software to other forms of cognitive production such as text, film, music, design, etc. (Benkler 2006; see also the P2P Foundation website). Wikipedia, in which thousands of individuals cooperate to produce an excellent free encyclopaedia, is a good example. Benkler (2006) compared Wikipedia with other encyclopaedias and showed that Wikipedia is comparable to Encyclopaedia Britannica, which is arguably the gold standard for encyclopaedias.

There have also been efforts made to organise the PP of hardware. Global Village Construction Set (GVCS) is an open-source community for the production of ecological friendly machines (see www.oepsourceecology.org.nyud.net/gvcs.php). In 2005, Adrian Bowyer and his collaborators launched an open-source project (RepRap) for the production of a three-dimensional printer that would be able to reproduce itself (Bowyer 2006). They produced the first example of such a printer (‘Darwin’) in 2008. The two subsequent generations of printer, ‘Huxley’ and ‘Mendel’, were produced in 2009 and 2010 respectively.

To what extent, then, can PP also be applied to material production, and what will be the particularities of material PP?

That digital technology already plays a hegemonic role in material production could prove instrumental for the expansion of material PP. Material production has two components: knowledge (R&D, design, software, etc.) and hardware. Whether automated or not, the knowledge component can be produced through PP in the Linux model. Therefore, the digital automation of the production of hardware can facilitate material PP. To the extent that hardware production can be digitally automated, the human contribution to production will be mainly knowledge, because raw materials are extracted from nature by machines, and machines are produced by machines run by knowledge and using raw material.

Therefore, the digitally automated PP of material goods is almost equal to the PP of knowledge. Digital technology already plays a hegemonic role in material production. The growing role of computer aided design (CAD), computer numerical control (CNC) machines and computer aided manufacturing (CAM) in material production are all evidence of this. This will facilitate the expansion of material PP.

Two caveats are in order here. First, the generalisation of automation need not mean the elimination of human physical activities. Actually, the automation of heavy and boring tasks may free human energy for more pleasurable physical activities. Indeed, it may provide a new basis for the revival of crafts. Second, not all material PP needs to be automated.

The most advanced form of digital hardware production currently is emerging is 3D printing. As we will see, the 3D printer (3DP), like PP, transcends both the social and technical divisions of labour. Therefore, the expansion of each will instigate the expansion of the other. Therefore, 3DP is the most optimal machine for PP.

What is 3D printing?

Since I deal more extensively with 3D printing elsewhere (Rigi, forthcoming), I will be very brief here. I concur with the Economist (2011) in that that 3DP may introduce a new industrial revolution. A 3D printer makes (prints) three-dimensional solid objects from digital files. It principally functions like an ordinary printer, except that its nozzle, which spreads the material onto the bed, can move in three dimensions.

The printer is guided by algorithms that extract data from computer-aided design (CAD), and accordingly lay down successive layers of material (powder, sheet, liquid, rod, etc.) atop each other until the object is produced. Ultra-thin layers of between 20-30 microns (0.02-0.03 mm) of material are spread on a bed, and are fused and solidified into the required patterns with a squirt of a liquid binder, or by sintering with a laser or an electron beam. The defining characteristic of 3DP is the additive manufacturing method: it is the opposite of the currently dominant subtractive manufacturing method, which operates through cutting, drilling, pressing, hammering and welding. 3D machines produce objects from various materials in an uninterrupted process in which forms are created according to the requirements of the functions. The process resembles pottery, with the difference that human organs (brain, hands and eyes) are replaced by the machine and by algorithms.

The most revolutionary characteristic of 3DP is the fact that it, like P2P, though in a different way, transcends both technical and social divisions of labour. The technical division of labour in capitalist enterprises constitutes the geometry of the labour process. It situates each moment of the labour process in a particular place. In 3D printing, this geometry is replaced with successive algorithmically guided movements of the printer’s nozzle. Hence, 3D printing holds the potential to abolish the factory regime altogether. Generalised 3D printing also abolishes the market. The social division of labour, which is the sum of different branches of production producing different goods, is also transcended. The 3DP is a generic machine, and can produce different objects from different material according to the algorithm used without being retooled (Economist 2011: 69-71). Hence, the expansion of 3D printing undermines the domain of commodity and market. Imagine a situation in which most household goods are printed at home with the aid of printers that are the size of a washing machine, and that can reproduce themselves. This holds the potential to bring the capitalist system to a decisive crisis. Generalised 3D printing could abolish altogether both the factory and the market. Individuals, households or small communities could download PP-produced designs from the net, customise them according to their own tastes and needs, and print them.

3D production, though still an emerging and marginal niche, is growing fast. In 2011, 20 per cent of all 3D products were finished objects (Economist 2011), and this grew to 28 per cent in 2012 (Economist 2012). Although three-dimensional printing was initially developed by commercial firms, the open-source movement has started projects on the open-source production of both 3D printers and 3D designs. Good examples of this are RepRap, as mentioned above and Thingiverse. Thingiverse is a project for the open-source production of digital design for three-dimensional objects.

The particularities of material PP

The materiality of the inputs and outputs of material PP gives particular properties to the processes of production and the distribution of goods. Whether fully automated or including human labour, the production of material components, by necessity, must happen in particular places. The material inputs are physically bounded and, in comparison with knowledge, scarce and must be procured. The product itself is also physically bounded, and located in the place of its production. The bounded nature of the process of material production means that the number of individuals cooperating in the material peer production of a particular good is necessarily limited. The transportation of the material input and output cost time and energy.

Unlike the digital reproduction of knowledge, the reproduction costs of material objects, everything else being equal, is the same as the cost of their production. Likewise, the relative scarcity and bounded materiality of products means that only a limited number of people at a given time can own and use them. Consequently, it is natural to assume that material PP will be a local affair. Moreover, it will be more convenient if members of producing communities live in the same locality. This will facilitate cooperation and the distribution of products. This will also minimise the costs of transportation and storage. Local communities can produce what they need in a just-in-time fashion.

It is important to note that these local communities of production cannot be self contained, but must be part of global networks in two ways. First, as mentioned above, knowledge is produced through global cooperation. Second, these communities need to procure some energy and raw materials from outside.

In a world in which the overwhelming majority of material production is done by 3DPs and other robots, all material production, except that of raw material and some types of energy, can be done locally. Primary production factors will consist of knowledge and nature (raw material and energy). Machines will reproduce themselves. But if knowledge is commonly produced and shared by all of humanity, then the major natural resources of nature must also be shared in the same way. The relative scarcity of such resources requires rules of just distribution. The limit of material wealth also requires rules for the just distribution of such wealth among members of each local community. But the criteria of distribution will not be the contribution of labour by individuals, because labour contribution is global, and has no exchange value. Thus, the criteria of distribution cannot be anything except need, as defined by the community.

A PP revolution? The role of social struggle

Indeed, an integrated expansion of PP and 3DP may push capitalism into a fatal crisis (see also Varcellone 2004), and ultimately subvert it. However, the victory of PP cannot be taken for granted. Capitalism will not collapse of its own accord. Everything will depend on social and political struggles, of which the expansion of PP is an important component. The final victory of PP, if it is to happen at all, will not be merely a result of the gradual expansion of the PP mode of production. It will also depend on social revolutions that abolish the private ownership of strategic natural resources and major infrastructure, and transform them into shared commons of humanity. Without such commons, PP will not become the dominant mode of production. Without such social revolutions, capitalism may succeed in subduing the threat posed by PP. State and corporate empires will fight tooth and nail to bring IT technologies under their full control, as they did with previous transformative technologies such as the radio (Wu 2010). However, the PP movement, if supported by all other anti-systemic social movements, different strata of working classes, and humanity at large, may prevail. There are good reasons to believe this.

First, the current epochal crisis of capitalist civilisation has revitalised the search for alternatives to capitalism (Douzinas and Žižek 2010). The growth of revolutionary and utopian imaginations and revolutionary energies provides a supportive intellectual and cultural milieu for the propagation of the idea of a PP revolution. Second, the historical agent that can lead the PP revolution is emerging in the bodies of anticapitalist protesters and PP producers. Both are components of what Hardt and Negri (2000) have called the multitude. The commonalities that follow may facilitate the forging of a strategic alliance between them. (1) Both anticapitalist protesters and PP producers are organised through voluntary horizontal cooperation. In this respect, they have developed common democratic procedures for accommodating participant views. (2) IT is essential for both. (3) A section of PP technological activists have played leading roles in the recent social protests in the West and the Arab world. (4) Anticapitalist protests both in the USA and Spain have embraced PP and the commons as an alternative to capitalism. (5) Anticapitalist movements and PP foster similar cultural values and ways of life which might be the seeds of a future collectivist culture. My fieldwork shows that there is a good cultural ground among PP communities for the reception of a radical critique of capitalism. The majority of hackers cherish collectivistic and progressive values. They are proud of producing commons instead of commodities, and have a strong commitment to the interests of their communities. They value democratic aspects of voluntary horizontal cooperation, and prefer creativity and happiness to money and careerism. They have concern for ecology, antipathy towards consumerism, and care for the poor and the Third World. They do not appreciate bragging, self-promotion, dishonesty and calculative manipulation (see also Weber 2008: 145-146).

Although these are important signs, a full alliance between these movements is still in the making. The formation of a solid collectivist and progressive culture that is an integral part of the PP movement will prove essential for the formation of a PP society. However, despite this progressive culture, the majority of hackers has no clear programmatic communist vision or sustained theoretical anti-systemic critique of capitalism. This makes PP vulnerable to invasions by capitalism. For instance, many projects that started as PP have been diverted into capitalist enterprises. Under this condition, the propagation of a clear communist vision among the contributors to PP production will be indispensable for its advancement. But this vision should emerge, and is emerging, from the theorisation and explication of PP and its potentials. The existence of a self-conscious communist section among PP producers is a good springboard for the production and spreading of this vision. The anticapitalist movements have not yet realised fully that PP can be an alternative to capitalism. Only some sectors of these movements, and only recently and not yet very clearly, have adopted commons and PP as alternatives to capitalism. A full convergence between PP and these movements will be essential for the final victory of PP over capitalism.

Third, as the PP alternative offers the labouring masses a way out of capitalist exploitation, crisis, insecurity and unemployment, it is in the interest of these masses to support and participate in PP revolutions. There is an objective basis for the alliance between knowledge workers, participants in PP and other type of workers in their struggle against capitalism. Capitalism exploits all of them, albeit in different ways.

Fourth, the majority of humanity will benefit from a PP revolution, and therefore it is in the interest of most human being to support it and participate in it. This is already evident from the current struggles around digital-cultural goods, and patent and copyright issues.

The idea that contradiction between productive forces and capitalistic relations of property can only be solved through a social revolution does not imply that social struggle has not been involved in shaping this contradiction. Marx showed (1976) that the struggle of wage labour for a shortening of the working day played a central role in the quest for production of relative surplus-value, which in its turn instigated successive technological revolutions. The social revolution is both a continuation of and a break with the daily struggle of wage labour. It is a continuation, because it builds on it; and it is a break, because it does not bargain for better conditions within the capitalist relations, but attempts to subvert them.

Conclusion

The contradiction between the de-territorialised nature of knowledge and IT and the territorialising characteristic of capitalist forms of property, and the division of labour and property represents a new stage in the inherent contradiction between the social nature of forces of production and private appropriation in the capitalist mode of production. PP is emerging as a response to this new stage of contradiction, and as a negation of capitalism. However, it is currently dominated by capitalism. Its advance and final victory over capitalism are functions of the struggle not only of knowledge workers, but of the multitude of all types of workers and social movements who strive for progressive revolutionary change.

Although PP is a historically new mode of production, it cannot by itself overthrow capitalism. Such an overthrow requires an alliance of revolutionary social movements. PP, however, can contribute to the formation of such movements by presenting a clear alternative to capitalism. The hackers’ movement can also become an important element of such a revolutionary alliance. Depending on the balance of forces in the realm of social struggle, the current peer production may evolve in any number of directions, including the following. (1) Capitalists colonise and use it as rent-extracting mechanism from non-knowledge workers within a predominantly capitalist context. (2) A new Keynesian state regulates its relation with a predominantly capitalist sector in order to boost a new type of the welfare state (Orsi 2009). (3) Cooperatives of knowledge workers control peer production within a predominantly capitalist context in order to extract rent from workers who work in the capitalist sector (Kleiner 2007). (4) Revolutionary social movements overthrow capitalism, abolish the state, transform the strategic resources into commons, and generalise peer production to all branches of production.