[p2p-research] What Networks Are (and What They Are Not) A Response to Galloway’s Position Paper

Tom Rawlings tom at fluffylogic.net
Fri Aug 6 19:14:13 CEST 2010


I've added this to the blog to come later this month...

In Alexander R. Galloway; Exploring New Configurations of Network 
Politics, he suggests that the fascinations with networks that has 
entered into so much critical thinking is not all it’s claimed.  That 
networks, by definition are asymmetric constructs and those using them 
as tool of understanding are failing to grasp this:

"It is common to talk about networks in terms of equality, that networks 
bring a sense of evenhandedness to affairs. It is common to say that 
networks consist of relationships between peers, and that networks 
standardize and homogenize these relationships. It is not important to 
say that such characterizations are false, but rather to suggest that 
they obscure the reality of the situation. Networks only exist in 
situations of asymmetry or incongruity. If not no network would be 
necessary– symmetrical pairs can “communicate,” but asymmetrical pairs 
must “network.” So in addressing the question “What can a network do?” 
it is important to look at what it means to be in a relationship of 
asymmetry, to be in a relationship of inequality, or a relationship of 
antagonism."

I’ve got to take issue with his initial definitions here:  Networks are 
constructions of two basic units – nodes and links.  The number of nodes 
and links is a very different matter to any power accrued by the node/s 
or bandwidth of the link/s.  Put simply Galloway is seeing 
power-relations in shadows and missing ones in plain sight.  Being a 
Peers in a network mean having the ability to form (and break) links – 
again it is not a reading of the links power/bandwidth.  If one node has 
many links, then it may lead us to conclude that it is a more powerful 
node than others, yet if that node is not a peer – i.e. if others can 
form links with it, but it cannot choose to deny them – then it’s power 
is illusory.  The same applies where two nodes are talking – what is to 
say they are a symmetrical pair?  Nothing.  In short, networks create 
tendencies – if you can create links, then the tendency is towards a 
less hierarchical structure.

We also see Galloway’s misinterpretation of tendencies as absolutes in 
his discussion of the Robustness Principe:

"The so-called “Robustness Principle,” which comes from RFC 761 on the 
transmission control protocol (TCP), one of the most important political 
principles of distributed networks, is stated as follows: “Be 
conservative in what you do, be liberal in what you accept from others.” 
This is called the Robustness Principle because if a technical system is 
liberal in what it accepts and conservative in what it does the 
technical system will be more robust over time. (But of course wouldn’t 
it ultimately make more sense to relabel this the Imperial Principle? Or 
even the Neoliberal Principle?) This indicates a second virtue of 
protocol: totality. As the Robustness Principle states, one must accept 
everything, no matter what source, sender, or destination."

The Robustness Principle is a tendency and not a command.  It does not 
force a node to accept everything.  Each node has the will have the 
capacity to interpret the principle differently – not as they are 
ordered to by any protocol.  Thus the parameters of operation are set by 
via each node, resulting is a a tendency towards non-hierarchical 
structures.  Protocols are methods of intercommunication – they are not 
structures of totality – for example, a node can use them to communicate 
with other nodes, while adhering to different protocols within its own 
communication.  Protocols are structures of common agreement underpinned 
with practical agreement – nodes who struggle to use a protocol will 
disengage from it, so reducing it’s use.  Whereas nodes that are 
conversant with a node will be taken up and used, so growing the 
network.  As such the action of the node is needed to make a protocol 
work – and as such it has a tendency towards a democratic method of 
operation.

Thanks

-- 
Tomas

-----------------------
Tomas Rawlings
Development Director, FluffyLogic Development Ltd.
web: www.fluffylogic.net
tel: 0117 9442233 
-
Also see:
blog on film & interweb: www.plugincinema.com
blog on games, p2p, media ecology & evolution: http://agreatbecoming.wordpress.com
tweet: www.twitter.com/arclightfire




More information about the p2presearch mailing list