Reviewer: Toni Prug, Queen Mary, University of London

Without knowing the insides of the FreeBSD, I can’t judge well the first part of the text. I do find it useful to learn from it, especially because I’m a firm believer that some of the practices from the software and networking communities would be very desirable in other walks of life.

pg.2. “Debian’s ‘rejection of market economy … in favour of cooperative production’ – (since I don’t have the complete section around the quote, this could be out of the content) it would be difficult to argue that Debian does not function fully within the capitalist market economies. Who pays their salaries (and/or their time to work on Debian) and what is the individual purpose and systemic role of those who pay them? These would more appropriate questions to determine the role of Debian in capitalist economies, rather then participants’ own views. Furthermore, it seems to me that a hugely disproportionate number of Debian developers (in comparison with overall society) has postgraduate education, while a significantly large number of them have good salaries in IT sector, academia or R&D. If this is true, then they belong to a tiny minority of a privileged class, which then raises a set of frequently overlooked questions on the class composition and class role of Debian, and more general of Free Software and Open Source projects.

More concretely, since the internal logic of capital for self-expansion creates inequalities and blocks the development of democratic participation both at the work place [1] and in political institutions [2], we have to examine what systemic role do the software and networking projects and communities play in the capitalism as a whole. Such analysis is necessary to acquire the knowledge that is good enough to judge the possibility of adopting, adjusted accordingly to different contexts, some of the their appealing cooperative practices across society. Otherwise, we may easily continue falling into trap (as many theorists and commentators do) of liking what we see in software and networking communities, but unaware of the limits that exist to have the appealing aspects of those forms be used across the society and unaware of how capital utilizes such pockets of different cooperation for strengthening of its rule.

Authority and legitimacy – Max Weber’s lens is a possible one for a microlevel of analysis, but it is a limited lens that fails the see the above critique of social relations in capitalism necessary for a macrolevel analysis.

For example: “As a general rule, FOSS projects ‘are created with few traditions to guide them and so do not inherit a traditional basis of authority’ (O’Mahony & Ferraro 2007, p. 1081). They do not rely upon a legal-rational basis of authority either, as there is no authoritative division of labour.” – this is true only on microlevel of software and networking projects in isolation. However, division of labour across society is the first and perhaps the most important feature of society that enables participants in those projects to do as they wish in their projects i.e. that is why they can focus only on what they do and not on everything else that makes our lives possible (food production, transport and energy infrastructure, education, health and other care institutions, common budget and the resulting spending which is a large part of the entire GDP, etc.). Furthermore, someone pays their wages and the time they spend on software produces value which is backed up by other value production, due to division of labour and uneven value distribution across society.

Another example why use of Weber is not sufficient (quotes inside the quote are Weber’s): “Direct democracy is characteristic of groups which, in order to preserve their members’ autonomy, attempt ‘to dispense with leadership altogether’ by reducing ‘to a minimum the control of some men over others’” – this definition of direct democracy is later used to describe FreeBSD participation and decision making.

When Weber writes about reducing ‘to a minimum the control of some men over others’ an accurate critical assessment of social relations in capitalism is entirely lacking. Since it is precisely the relationship of capital-labour that determines, far more than anything else, the control over some people over other, thus placing limits on possible forms democratic cooperation can take.

In Weber, there’s minimal, if any at all, place for such, most important critique of the conditions of possibility for a directly democratic society. In short, those who are forced to sell their labour the longest, for lowest wage, under worst labour conditions, have the worst chances of directly democratic participation – regardless of what mechanisms are in place.

When we add the time an average worker in a highly developed region like USA, or Europe, sells labour and commutes on daily basis, plus the unavoidable administration of life and time necessary to spend with the family, there is hardly any time left for participation in anything that we might meaningfully call direct democracy (or democracy at all) – wage labour consumes it all.

Yet, because such critical assessment is widely known and available to social scientist, primarily, but not only, through Marx and Marxist authors, using Weber for a description and evaluation of democracy (direct or otherwise), is a seriously significant deficiency.

Equally important, Weber’s own understanding of democracy was a lot more nuanced and ambiguous then what we get from this text. Weber puts emphasis on the autonomy of authority and bureaucracy over the democratic elements, and puts democracy in quotation marks in several places (in the The Theory of Social and Economics Organization used by the author) questioning its meaning.

Talcott Parsons acknowledges this in the introduction to the book:

“In the first place he [Weber] calls attention to the fact that two of the most important types of check on centralized authority, the separation of powers and the presence of collegial bodies in place of monocratic positions of authority, are primarily associated with aristocratic rather than democratic regimes.”

The history of English parliament have had some form of representation, collective bodies and separation of powers many centuries before political parties came to existence, centuries before the system was developed further, tweaked and renamed to be “democratic” because inhabitants select an interest group (political party) over another for a four or five year long periods of almost entirely unaccountable rule. The problem is clearly visible in the conclusion of the text too:

“If, in the Weberian tradition, we take the basis of authority as the decisive organisational feature, then the mode of organisation of FreeBSD is collectivist, based on direct-democratic procedures of decision making.”

Weber’s work, even at its ambiguous best, is insufficient for evaluating democratic or collectivist mode of organization. Given the wealth of information and initial framing of the internal functioning and problems of the FreeBSD project, it is regretful that analytical framework selected is so insufficient and lacking.

[1] through the concept of private property specific to capitalism, democracy at workplace is denied and ruled out as impossible

[2] interests of capital to keep forms of strictly representative political system through which lobbying, donations, private ownership of mass media and personal wealth determine who ends up in parliaments, and interest to keep extracting surplus value in form of profits from long working hours, block the development of direct democracy, of wide participation in discussions and decision making on common affairs, across society.

On to the formal questions:

1) Is the subject matter relevant?

Yes

2) Is the treatment of the subject matter intellectually interesting? Are there citations of bodies of literature you think are essential to which the author has not referred?

Intellectually interesting? Yes, at first, disappointing when the analysis arrives. Author has referred to Weber who is, in the way author used him, entirely insufficient for the analysis. A use of more ambiguous Weber’s, who doubts what democracy is, would have been less problematic, since it would have left the issue of democracy open. However, it is an overall arrangement in society that enables the privileged mode of cooperation which author has set to analyse – not only democracy, but the question of authority and charisma cannot be sufficiently analysed without such, overall, analytical framework. Essential body of literature for this? Yes, Marx and Marxists, and heterodox economists too – software and networks have been essential for the explosion of financialization and neo-liberalism in the last several decades. While inequalities have vastly risen across Western states in those decades, analysed software and networking communities have gained special labour conditions, choosing even own mode of cooperation. Microanalysis of individual projects, or even whole sectors, developed without macro conditions of society at large are easily prone to not only giving wrong answers, but more importantly so, to posing the wrong, or not good enough, questions in the first place.

3) Are there any noticeable problems with the author’s mean of validating assumptions or making judgements?

As it is, the concept of democracy in the text is entirely devoid of time (consumed by wage-labour), space (dwelling. Or, in modern terns, a mortgage that ties most of labour force further to the necessity to sell own labour for a wage devoid of any democratic elements in the way and purpose own labour and its results gets used), and other forms of class analysis (privilege of education, cultural background, social circles one belongs to, family wealth). Yes, given the centrality of the concept of democracy for the text, there’s a big problem with Weber per se , and furthermore with author’s use of Weber – see above for more details.

4) Is the article well written?

Yes.

5) Are there portions of the article that you recommend to be shortened, excised or expanded?

Using Weber’s more ambiguous thoughts on democracy would be an improvement. Removing, or drastically toning down, claims on what is, what is not, and why, a democratic form of organization would be a significant improvement.

Reviewer: Felix Stalder, Zurich University of the Arts

Review of The Emergence of Governance in the FreeBSD Project

This is a very good paper, well researched, well written and addressing an issue – governance is peer production — that is timely and important. The combination of a longitudinal case study and broader theoretical perspectives is very productive, particularly in relation to the core concern of the paper, the transformation of one mode of governance into another, with both differences and continuations between them.

My comments should be seen as an engagement with the paper, rather than some direct requests for revisions.

First, it remains somewhat unclear where the author relies on secondary research and where the paper is based on original research, which is hinted at in the research methodology section. Some clarifications would be helpful.

Second, on page 11, final paragraph, the author mentions the classic trade-off between flexibility of small-scale, information organizations and the introduction of formalized hie rarchies to manage largescale processes. He goes on to argue that the project managed to avoid this trade-off by standardizing many of its procedures, particularly around the inclusion of new committers. I think this is an insufficient explanation, given the centrality of the problem. Manuel Castells, for example, made the dissolution of this trade-off through network-technologies one of the central element of this theory of the network society. He argued that due to the increased information processing capacities, organizations can now be large and flexible. Perhaps some references to this discussion might help to flesh out why the scaling issue was less of a problem for the project than what one might expect and why relatively little changes to the organization were sufficient to manage it.

Third, the strong reference to Weber is both productive and limiting. In the end, the author points out that “none of Weber’s categories captures sufficiently the character of authority in FreeBSD.” This would suggest that there is need to develop Weber’s categories further. The author already hints it it, when he calls the community an “informal meritocracy” but does not develop it further. Perhaps expanding on the role of meritocracy might help to overcome a difficulty that the current argument does not resolve. In many ways, in open source projects, authority is based on “rational charisma” which, in the Weberian perspective, is a contradiction in terms, since charismatic authority is, per definition, irrational. But in a functioning meritocracy, rationality and charisma might not contradict each other. Of course, calling open source governance meritocratic is not particularly original, but strength of the paper lies in being able to show how the meritocracy is made to actually work in a particular context.

Finally, the paper would profit by re-connecting its findings the larger field. The literature review in the beginning is well done, but in the conclusion it would be helpful to revisit it, in order to highlight the particular contribution – by confirming or questioning previous explanations – the paper makes to the broader debate.

A minor remark, Mateos-Garcia is sometimes incorrectly referenced in text as “Garcia”.

Reviewer: Ed Steinmueller, University of Sussex

1) Is the subject matter relevant?

Yes.

2) Is the treatment of the subject matter intellectually interesting? Are there citations of bodies of literature you think are essential to which the author has not referred?

Yes.

3) Are there any noticeable problems with the author’s mean of validating assumptions or making judgements?

No.

4) Is the article well written?

Yes.

5) Are there portions of the article that you recommend to be shortened, excised or expanded?

See more detailed comments following.

Larger Issues

Weber is now introduced early but without an explanation of how his ideas will be applied and the reader must wait until p. 15 to learn how Weber is employed. The discussion on p. 15 to the second complete paragraph of p. 16 belongs at the beginning after Weber is introduced. Ideally, the descriptive account would then make reference to the application of Weber as it is developed which would make it possible for the current material from the final paragraph at p. 16 which continues on p. 17 and through the complete paragraph on p. 18. The specific discussion of the role of charisma is appropriately placed as a refrain and further development in its current location where it can also contribute to the conclusions of the paper.

The discussion at p. 14 belongs earlier to set the context for the discussion of committer autonomy. It provide a much better motivation for the earlier descriptive material on the evolution of governance and the ‘just the facts’ approach to that material which risks putting off readers who may be impatient with unmotivated descriptive narrative (even as well written as this narrative).

In the conclusion, perhaps it is not only Weber that should be found to have shortcomings in explaining the operation of the BSD community but also Mintzberg.

Smaller Issues

p. 9 Table 1 and the text discussion are needlessly duplicative.

p. 11 The episode recounted at the paragraph beginning at p.10 and continuing on p.11 leave unresolved the larger question of the fate of the suspended committer in the elections, a natural question for the reader.
p. 11 The discussion of forking emphasises their role in ‘mitigating’ and ‘dampening’ conflict which can be read in two different ways. One reading would be the ‘knives on the table’ nature of forking which emphasises that the possibility of forking is essentially a deterrent threat which suppresses the expression of conflict. The other reading is that the possibility for exit represented by forking actually reduces conflict by leading to spin off of dissident groups, thus depleting potential sources of conflict. In principle, this is an empirical question. In several communities that I have examined, it appears that the former is the better explanation. The ambiguity is worth resolving for the case at hand. Forking such a large project is problematic since recruitment of the human resources to pursue the alternative path will not be straightforward and thus I would expect the former to apply here as well.

p. 12 The discussion of the paragraph beginning on p. 11 and continuing on p. 12 is somewhat duplicative of the preceding discussion at p. 5 in which a more descriptive account of the mentorship phase of the new committer’s life cycle is presented. Particularly the sentence that begins ‘Usually, the committer vouches for …” A minor touch up here will reduce the reader’s déjà vu experience.

p.12 In addition to the Cusumano and Selby and Mintzberg references it might be useful to consider the similarity of this process with Japanese manufacturing technique in which individual workers are given authority to stop production lines when defects are identified, thus dramatically reducing the complications resulting from production of defective final products that have to be ‘reworked’ – a typical reference would be Schonberger, R. J. (1982). Japanese Manufacturing Techniques: Nine Hidden Lessons in Simplicity. New York, The Free Press.

Thank you for the feedback.

The first reviewer requests changes that are relevant to the
subject-matter of the paper but I am afraid that some of the most
important concerns he raises will have to be more fully addressed in a
follow-up paper in the future. I will explain why. He calls attention
to two main shortcomings of the paper. The first is that the paper
disregards the fact that FOSS development is embedded in capitalist
social relations. This has serious analytical implications: the paper
is censured for being one-sided: it explores the role of agency (the
subjective conditions of FOSS development) but overlooks that of
structure (the objective conditions). I agree that this is a valid
criticism. My goal in writing this paper was not to explore the uneasy
relationship between FOSS projects and the capitalist system embedding
them (that I did in a paper I wrote back in 2009 with Johan Söderberg
entitled ‘The hacker movement as a continuation of labour struggle’,
which was published in issue #97 of Capital & Class) but to test the
organisation theory which holds that the split of a group into those
who make decisions and those who execute them is a necessity imposed
upon the group by such structural changes as increased size. In
addition, the level of analysis I used was dictated by the
subject-matter itself: the paper explores the articulation of
authority within a FOSS project, not the conditions in society (such
as free time, education, etc.) enabling the FOSS phenomenon. However,
to address the concerns raised by the reviewer I added a paragraph at
the end which discusses the economic conditions underlying peer
production.

The second criticism he raises is that my employment of Weber is
inappropriate. Well, that’s a long discussion, but the short answer is
I disagree. Although it is not spelled out in the text, I used Weber’s
concepts in a manner that is identical with Weber’s ideal-typical
approach: I construct the concept by specifying the characteristics I
find most important to it and then I contrast this ‘ideal-type’ with
the empirical case. My employment of Weber is not that of a theorist
of (direct) democracy; and if I have adopted his definition of direct
democracy, that is because I find it a good working definition of the
concept of direct democracy, rather than due to an uncritical
absorption of his writings. This paper, in other words, does not aim
to comment upon Weber’s views on direct democracy (though that would
certainly make for an interesting paper and in fact it is one I’m
currently working on).

The second reviewer makes four comments. First, it is not clear what
in the paper is based on secondary research and what not. It is true:
I could have said somewhere that with the exception of the numbers
given in the text (e.g. number of developers with commit rights,
codebase growth, etc.) all other information – but not the
interpretation of that information, of course – is based on things
already published on the net or elsewhere. I will make sure to add a
few lines in the methodology section to clarify it.

Second, he suggests an analysis of the observed outcome (i.e. the
specific course of action taken by the FreeBSD project to cope with
increased scale) from a different perspective. This finding, writes
the reviewer, must be interpreted in the context of a prominent
management discourse which claims that networking technology enables
organisations to be large and flexible at the same time. There is no
doubt an element of truth in this – and I will make sure to
incorporate this point in the final version of the paper – but I chose
not to emphasise the role of technology more than that of the desires
of FOSS developers and of the normative standard (the hacker ethic)
governing their conduct, because it is the latter I find more
important in explaining the outcome.

Third, he suggests that the concept of meritocracy be further
explored. The reason I didn’t explore the concept of meritocracy any
further is because I couldn’t see a way to go any further with it.
What I did is I analysed the conception of merit in the project and
how it has changed (or not) over time. But I intentionally didn’t try
to formulate a general theory of what a meritocratic organisation is
or what it would look like. So, when I say in the text that FreeBSD is
a meritocratic organisation, what I mean is that administrative
authority in FreeBSD is distributed in accordance with the conception
of merit in the project. It is in this sense that I refer to FreeBSD
as meritocratic. But that should be obvious from the overall
discussion in the text. Then again, perhaps I should have been more
clear about this in the text and add a line to clarify it.

Four, the second reviewer suggests that the findings be framed in the
context of the literature review. I think they are. Maybe not in the
concluding section, but the analysis continuously reflects on the
similarities and differences between FreeBSD and what others have
found in their studies of Debian. Of course, this could have been done
in the final section of the paper as well, but I considered it’d be
more fitting to close the paper by summing up the key features of
FreeBSD’s governance, rather than how it differs from other FOSS
projects (as my goal was not to arrive at a typology of different
modes of FOSS governance).

Again, I appreciate the feedback and will try to address the points
raised by the reviewers in the revised version.