The challenge is not P2P but democracy and accountability

Bill Thompson
1 July 2003

In his essay Siva Vaidhyanathan works hard to establish the cultural and political significance of the growing number of peer-to-peer (P2P) networking tools that have come into general use on the Internet. He sees an opposition between the two political positions of anarchy and oligarchy, an opposition which extends from the real world onto the Net, and outlines a dialectic between the positions which derives from the widespread adoption of P2P. This, he contends, is ‘remaking our information ecosystem’ and altering the world.

Unfortunately, while I think his core thesis is both interesting and defensible, the approach he has chosen simply does not support it adequately, and the central argument is both weak and technologically flawed. While the view that ‘anarchistic’ communication technologies are particularly important where there is no ‘rich, healthy public sphere’ – we should encourage email, blogging and chat in Myanmar, Iran and Afghanistan – has much to recommend it, Siva’s essay would fail to convince a sceptic even if it entertains those who already agree with him.

P2P: protocols or architecture?

As a former programmer who still writes the occasional programme in C and PHP, I am particularly concerned with Siva’s apparent lack of understanding of the technology of peer-to-peer networks. In his essay he seriously misrepresents both the nature of P2P technologies and their patterns of use. The result is that nobody who is familiar with the history, use and technical architecture of today’s P2P systems will find his wider argument at all convincing.

The central problem is a confusion between a specific type of P2P system, those like Gnutella and KaZaA that are used for file sharing over the Internet, and the general field of peer to peer networks. When Siva states that “P2P technology is a set of protocols that allows users to open up part of their private content to public inspection and thus copying” he is not merely oversimplifying, he is wrong.

First, because ‘peer to peer’ is not a set of protocols but an architecture for the development of information systems based around the core idea that both programmes in a network connection can have equal status – be ‘peers’ – instead of having a hierarchy in which one, the client, makes a request and the other, the server, satisfies it. Second, because only a small proportion of the large and growing number of P2P applications are concerned with file sharing, which is what he is talking about here. Others are concerned with collaborative publishing, software development, data validation and community services.

He is also wrong to say that you can only access a file by making a copy of it. Streaming technologies like RealPlayer and Windows Media Player allow access/viewing without copying (except in the trivial sense that the bits are moving over the network to your computer and appear in your frame buffer). Record companies and broadcasters make extensive use of this facility – you can, for example, listen to the BBC World Service on your PC but are not permitted to save the programmes you hear to a disk for later playback. You can get software to do this, but it works by grabbing the bits and is not part of the normal operation of the audio playback software.

The later claim that P2P systems try to ‘simulate the structure and function of the original Internet’ is similarly inaccurate. Siva is correct when he points out that the original Internet allocated fixed IP addresses to every node, so that each could participate as a server as well as a client. He is also right that the rise of dialup networks and dynamic addressing messed this up to some extent – but far more significant were the decisions by ISPs to introduce differential bandwidth so that downloading was faster than uploading, and of large corporations to put their LANs behind firewalls and use network address translation (NAT) to hide their internal addresses from potential hackers.

P2P anonymity is a sham

Siva’s claim that ‘most of what happens over peer-to-peer networks is relatively anonymous’ is similarly misleading. It may go untracked, but that does not imply anonymity in any strong – or legally useful – sense. With the most popular P2P file sharing tools each node is identified to the network and when logged on its IP address is known. If a record company sets up its own KaZaA node then it can easily grab IP addresses in real time and, if it was working with local law enforcement, probably come round knocking at people’s doors while they were in the act of uploading files.

US ISP Verizon recently handed over the names and addresses of four of its customers to the Recording Industry Association of America, demonstrating clearly that the supposed anonymity is a sham.

People who do not understand the technology think it is anonymous, but US ISP Verizon recently handed over the names and addresses of four of its customers to the Recording Industry Association of America, demonstrating clearly that the supposed anonymity is a sham.

It is also wrong to claim that P2P file sharing networks are unmediated and uncensorable, at least not as currently implemented. For one thing the network necessarily mediates the communications taking place over it. If Siva means that the record companies are out of the loop when it comes to music distribution then he should say it clearly, and not attempt a philosophical sleight of hand: unmediated it carries with it a weight of meaning that the practical reality of Napster cannot bear. It is equally foolish to claim that any part of the Net is uncensorable: any communication over the network can be tracked and censored, and even if I can’t read the content I can block the packets. China and Saudi Arabia censor the Internet, even file-sharing networks, very effectively this way. The meme that “the Internet sees censorship as damage and routes around it” is an effective soundbite but does not reflect the technical reality of today’s network.

Finally, while an analogy between P2P-based conversation and gossip is useful, it only applies to the social software P2P systems, not to Gnutella, KaZaA and the rest of the file sharing programmes. Gnutella is not a conversation, it is just a trading system – I do not chat to or interact with the people at their computers, I just search and copy files from them.

The analogies don’t stand up

This is a long list, and it may seem like I am quibbling about detail. But the effect of these many failures to understand just what P2P networks are, how they work and what they do is to undermine the core political argument that “peer-to-peer models of communication have grown in influence and altered the terms of exchange”.

This becomes most clear in the section ‘Listening to Napster’ when we are expected to agree that “certain sections of modern society have evolved with and through the ideology of peer-to-peer”. The analogies between peer-based social networks and P2P software simply do not stand up. Peer review and peer-to-peer are not the same thing at all, and peer production of open source software is not P2P production. Siva is exploiting the fact that the same word is used to confuse matters so that he can later claim for P2P far more social and political significance than it actually merits.

We need to be clear here: in the context of P2P the ‘peer’ is a programme running on an Internet-connected computer, able both to initiate and accept network connections and to act both as a client and a server. This is not, in itself, special, radical or transformative. The ‘peers’ in academic discourse and computer programming are people – the ‘peers’ in P2P are not. I do not think our understanding is enhanced if they are confused or elided.

Good governments are possible

I also have difficulties with the political argument put forward here. Presenting the current debate as a battle between anarchy and oligarchy is overstating things, and ignores the fact that democratic governments may legitimately want to regulate both the market and the behaviour of consumers/citizens – sitting on both sides of the fence, passing laws which guarantee civil rights and which constrain monopolies, or backing a new copyright bargain which limits what the record companies can do. The choice is not solely between West Coast ‘freedom’ and bad governments/bad corporations: good governments are possible and, I would argue, desirable.

If we look at the way the Net has grown, and its likely future development, then we can see that the choice facing us is not about anarchy versus oligarchy, because the Net has never really been outside government and corporate control. Even at the beginning, governments set the parameters for the ARPAnet’s development, and today corporations decide what we can and cannot do online. Anarchy has never been an option, and P2P file sharing does not make it one today.

The choice

There is a choice to be made, although it is not about whether or not we have a governed and regulated Internet.

If we want an open network then those guarantees have to extend to the technology which provides and sustains the network, and they can only be extended by the state.

It is about whether the network is open or closed, whether it is run by the corporations in their interests or by democratically accountable governments in the interests of us all. An open society is neither unregulated nor anarchistic, but relies on state-backed guarantees of freedom in markets, property rights, freedom of speech, academic freedom and freedom from intrusive surveillance or loss of privacy. If we want an open network then those guarantees have to extend to the technology which provides and sustains the network, and they can only be extended by the state.

There is definitely a need for ‘healthy public discussion’ here, but this essay does not contribute much to it in its present form. Freenet is not a realistic solution – democratic control of a trusted network and accountable systems is the answer, not a techno-anarchism that can only fail to hold against the corporations.

In the end, Siva’s thesis is entertaining but simplistic: putting the US government and the People’s Republic of China on the same side in some claimed ‘infowar’ may sound radical but is really just misleading. And the idea that ‘anarchistic’ communication is a tool to undermine closed societies smacks of the worst sort of US interventionism. Is Saudi Arabia to be hacked because the social norms there make looking at photographs of women in bikinis impermissible? Should UK anti-war websites be undermined because they do not support the US imperial worldview?

Had enough of ‘alternative facts’? openDemocracy is different Join the conversation: get our weekly email


We encourage anyone to comment, please consult the oD commenting guidelines if you have any questions.
Audio available Bookmark Check Language Close Comments Download Facebook Link Email Newsletter Newsletter Play Print Share Twitter Youtube Search Instagram WhatsApp yourData