Has anyone else on this list been struck by the parallel between the current portrait of de-centralized terrorist networks emerging in the media coverage and the organizational logic of distributed computer networks? Certainly both forms of networking emerged with a similar goal in mind: a resiliance to the forms of centralized large-scale attacks characteristic of warfare in the first half of the 20th century. Redundancy and de-centralization are defensive structures whose effectiveness is demonstrated by the fact (reported yesterday, I think) that despite the destruction of something like 10 percent of Manhattan's office space, suprisingly little data was lost. In the face of this kind of distributed networking, the type of military response envisioned by Bush/Cheney et alia seems disturbingly out of joint. Just as we wouldn't imagine that we could take down a network by hitting a node, so too does the goal of "taking out" Osama bin Laden seem more symbolic than effective (to the effect that it works to proliferate cells of resistance, it might even be read as counter-productive). I'm wondering if there's some way to use the commonly accepted discourse on computer networking to shed some light on the current debate over the appropriate U.S. response to the recent acts of terrorism.
This is right on. I would also suggest that this attack will probably accelerate calls for organizations to become distributed networks with regard to personnel as well with regard to computers--who wants to come in to a central office location when such are a more attractive target. These accelarated calls should also accelerate technology to make such personnel distribution possible--video-conferencing technology and the like. --Christian Nelson "Prof. Andrejevic" wrote:
Has anyone else on this list been struck by the parallel between the current portrait of de-centralized terrorist networks emerging in the media coverage and the organizational logic of distributed computer networks? Certainly both forms of networking emerged with a similar goal in mind: a resiliance to the forms of centralized large-scale attacks characteristic of warfare in the first half of the 20th century. Redundancy and de-centralization are defensive structures whose effectiveness is demonstrated by the fact (reported yesterday, I think) that despite the destruction of something like 10 percent of Manhattan's office space, suprisingly little data was lost. In the face of this kind of distributed networking, the type of military response envisioned by Bush/Cheney et alia seems disturbingly out of joint. Just as we wouldn't imagine that we could take down a network by hitting a node, so too does the goal of "taking out" Osama bin Laden seem more symbolic than effective (to the effect that it works to proliferate cells of resistance, it might even be read as counter-productive). I'm wondering if there's some way to use the commonly accepted discourse on computer networking to shed some light on the current debate over the appropriate U.S. response to the recent acts of terrorism.
_______________________________________________ Air-l mailing list Air-l@aoir.org http://www.aoir.org/mailman/listinfo/air-l
-- NetscapeCommunicator4.73Win3220000502
At the risk of taking this somewhat off-topic, I'd wonder whether operating in this fashion would, in fact, make an organization more vulnerable to discovery. The notion of distribution via Internetworking, to my knowledge, involves concepts like store-and-forward, copies of information shared at all nodes, etc. One of the problems with "early" (in quotes because it wasn't that long ago) forms of Internet legislation in the U.S. was that it wasn't clear whether all of the nodes through which a pornographic image, for example, were equally liable under the law, because all of them would, by virtue of having a copy of the image, be "responsible" for its distribution. Would it be desirable for a group seeking secrecy to operate in such a fashion? It seems to me to run counter to the level of secrecy one would want, and that it would be better to have nodes that were actually independent of one another, or at least not very well informed of the others, so that if one _is_ taken down the others to which it might in some way be connected are not so easily discovered. Perhaps the notion of "network" as it's being applied here isn't quite the same as we've understood it in terms of the Net, or I'm misunderstanding its use in one or both of these contexts, or it's being used metaphorically? As to organizations like the ones housed in the WTC, Christian makes an interesting comment. I suspect most of the ones that were using computer databases had multiple backups, some of which were well beyond the WTC, possibly even in other countries. Why not similarly distribute personnel? And if I may be allowed a bit of tongue-in-cheek, perhaps "distance learning" could in this way come to mean sending faculty to teach from Hawaii while their students remain in Chicago. Sj
This is right on. I would also suggest that this attack will probably accelerate calls for organizations to become distributed networks with regard to personnel as well with regard to computers--who wants to come in to a central office location when such are a more attractive target. These accelarated calls should also accelerate technology to make such personnel distribution possible--video-conferencing technology and the like. --Christian Nelson
"Prof. Andrejevic" wrote:
Has anyone else on this list been struck by the parallel between the current portrait of de-centralized terrorist networks emerging in the media coverage and the organizational logic of distributed computer networks? Certainly both forms of networking emerged with a similar goal in mind: a resiliance to the forms of centralized large-scale attacks characteristic of warfare in the first half of the 20th century. Redundancy and de-centralization are defensive structures whose effectiveness is demonstrated by the fact (reported yesterday, I think) that despite the destruction of something like 10 percent of Manhattan's office space, suprisingly little data was lost. In the face of this kind of distributed networking, the type of military response envisioned by Bush/Cheney et alia seems disturbingly out of joint. Just as we wouldn't imagine that we could take down a network by hitting a node, so too does the goal of "taking out" Osama bin Laden seem more symbolic than effective (to the effect that it works to proliferate cells of resistance, it might even be read as counter-productive). I'm wondering if there's some way to use the commonly accepted discourse on computer networking to shed some light on the current debate over the appropriate U.S. response to the recent acts of terrorism.
_______________________________________________ Air-l mailing list Air-l@aoir.org http://www.aoir.org/mailman/listinfo/air-l
-- NetscapeCommunicator4.73Win3220000502
_______________________________________________ Air-l mailing list Air-l@aoir.org http://www.aoir.org/mailman/listinfo/air-l
Regarding the net as a metaphor/model for terrorist and other distributed organizations like vacationing teachers, it was notable today that one sector that apparently did well in a bad day for stock trading was videoconferencing technology. People apparently want to stay put.
As to organizations like the ones housed in the WTC, Christian makes an interesting comment. I suspect most of the ones that were using computer databases had multiple backups, some of which were well beyond the WTC, possibly even in other countries. Why not similarly distribute personnel?
And if I may be allowed a bit of tongue-in-cheek, perhaps "distance learning" could in this way come to mean sending faculty to teach from Hawaii while their students remain in Chicago.
Sj
Nancy Baym, Communication Studies University of Kansas NEW! email: nbaym@ku.edu NEW! snail mail: 102 Bailey, 1440 Jayhawk Blvd., Lawrence, KS 66045, USA NEW! url: http://www.ku.edu/home/nbaym
Nancy Baym wrote:
Regarding the net as a metaphor/model for terrorist and other distributed organizations like vacationing teachers, it was notable today that one sector that apparently did well in a bad day for stock trading was videoconferencing technology. People apparently want to stay put.
I guess some other folks had the same idea I had. Wish I'd had their kind of money to invest with! Per my response to Steve, I guess I ought to be investing in scenic real estate right now, too. Oh well. --Christian Nelson
Steve Jones wrote:
At the risk of taking this somewhat off-topic, I'd wonder whether operating in this fashion would, in fact, make an organization more vulnerable to discovery. The notion of distribution via Internetworking, to my knowledge, involves concepts like store-and-forward, copies of information shared at all nodes, etc.
Certainly, my notion of "distributed network," as applied to an organization's personnel, didn't presume what was just described. But it seems like people do use the term "distributed network," at times, to refer just to networked computers which are simply interconnected in such a way that there is no centralized site of control--i.e., a network that can't go down simply because one site goes down. And it seems that this is, or has often been claimed to be, the raison d'etre of the Internet (or at least its predecessor, Arpanet). Perhaps the meaning of the term "distributed network" has evolved, or perhaps the term has just developed more than one sense.
One of the problems with "early" (in quotes because it wasn't that long ago) forms of Internet legislation in the U.S. was that it wasn't clear whether all of the nodes through which a pornographic image, for example, were equally liable under the law, because all of them would, by virtue of having a copy of the image, be "responsible" for its distribution. Would it be desirable for a group seeking secrecy to operate in such a fashion? It seems to me to run counter to the level of secrecy one would want, and that it would be better to have nodes that were actually independent of one another, or at least not very well informed of the others, so that if one _is_ taken down the others to which it might in some way be connected are not so easily discovered.
I'd agree.
As to organizations like the ones housed in the WTC, Christian makes an interesting comment. I suspect most of the ones that were using computer databases had multiple backups, some of which were well beyond the WTC, possibly even in other countries. Why not similarly distribute personnel?
In fact, I was just talking to someone whose company, Mass Mutual, took up a fair amount of space in the WTC. He says that that office downloaded all their stuff off-site at the end of every day.
And if I may be allowed a bit of tongue-in-cheek, perhaps "distance learning" could in this way come to mean sending faculty to teach from Hawaii while their students remain in Chicago.
While this is tongue-in-cheek it does point to a problem with the potential trend of organizational distribution of personnel--more and more folks, freed from geographical constraint, will head toward the most scenic areas of the world and despoil them by overrunning them. --Christian Nelson
At 10:07 AM -0400 9/18/01, Christian Nelson wrote:
Steve Jones wrote:
At the risk of taking this somewhat off-topic, I'd wonder whether operating in this fashion would, in fact, make an organization more vulnerable to discovery. The notion of distribution via Internetworking, to my knowledge, involves concepts like store-and-forward, copies of information shared at all nodes, etc.
Certainly, my notion of "distributed network," as applied to an organization's personnel, didn't presume what was just described. But it seems like people do use the term "distributed network," at times, to refer just to networked computers which are simply interconnected in such a way that there is no centralized site of control--i.e., a network that can't go down simply because one site goes down. And it seems that this is, or has often been claimed to be, the raison d'etre of the Internet (or at least its predecessor, Arpanet). Perhaps the meaning of the term "distributed network" has evolved, or perhaps the term has just developed more than one sense.
Perhaps we can separate out two things from this. One is the notion of a network as a means of transmission, of moving information from one place to another. That notion certainly allows for the use of "distributed network" in terms above. If one part of the network goes down the rest of the network can communicate. But we can also use the notion of a network as a means of storage and retrieval. In those terms, if a part of the network goes down, the other parts would have to have "mirrors" of content. If nothing else it's an interesting way to think, again, about the differences between "transmission" and "ritual" views of communication. If the network is considered one of transmission, distributed networking makes sense as a means of ensuring communication is maintained if parts of the network are vulnerable. But if the network is considered one of information sharing, memory, etc., a loss of a part of the network is a loss to the whole network unless each part is "backed up" at every other part.
While this is tongue-in-cheek it does point to a problem with the potential trend of organizational distribution of personnel--more and more folks, freed from geographical constraint, will head toward the most scenic areas of the world and despoil them by overrunning them.
Although I don't discount that possibility and would be dismayed by its realization, I'd point out that it seems like most scenic areas of the world have been despoiled and overrun by the organized distribution of personnel for purposes of recreation. Sj
Steve Jones wrote:
At 10:07 AM -0400 9/18/01, Christian Nelson wrote:
Steve Jones wrote:
At the risk of taking this somewhat off-topic, I'd wonder whether operating in this fashion would, in fact, make an organization more vulnerable to discovery. The notion of distribution via Internetworking, to my knowledge, involves concepts like store-and-forward, copies of information shared at all nodes, etc.
Certainly, my notion of "distributed network," as applied to an organization's personnel, didn't presume what was just described. But it seems like people do use the term "distributed network," at times, to refer just to networked computers which are simply interconnected in such a way that there is no centralized site of control--i.e., a network that can't go down simply because one site goes down. And it seems that this is, or has often been claimed to be, the raison d'etre of the Internet (or at least its predecessor, Arpanet). Perhaps the meaning of the term "distributed network" has evolved, or perhaps the term has just developed more than one sense.
Perhaps we can separate out two things from this. One is the notion of a network as a means of transmission, of moving information from one place to another. That notion certainly allows for the use of "distributed network" in terms above. If one part of the network goes down the rest of the network can communicate. But we can also use the notion of a network as a means of storage and retrieval. In those terms, if a part of the network goes down, the other parts would have to have "mirrors" of content. If nothing else it's an interesting way to think, again, about the differences between "transmission" and "ritual" views of communication. If the network is considered one of transmission, distributed networking makes sense as a means of ensuring communication is maintained if parts of the network are vulnerable. But if the network is considered one of information sharing, memory, etc., a loss of a part of the network is a loss to the whole network unless each part is "backed up" at every other part.
Good point re: the views of communication.
While this is tongue-in-cheek it does point to a problem with the potential trend of organizational distribution of personnel--more and more folks, freed from geographical constraint, will head toward the most scenic areas of the world and despoil them by overrunning them.
Although I don't discount that possibility and would be dismayed by its realization, I'd point out that it seems like most scenic areas of the world have been despoiled and overrun by the organized distribution of personnel for purposes of recreation.
Good point. Many have been so despoiled, but not all, and that's all I have to say about that, lest my favorite unspoiled places become the next popular place to trample. I should note that the difference between tourism and the type of population distribution I'm talking about is that it may not despoil so much as make the beautiful inaccessible due to the privization of right-of-ways, the development of former camp grounds, etc. --Christian Nelson
Steve's focus on the storage and retrieval aspects of a network are particularly interesting to me and I would like to toss out a few comments. In neural networks, it is the network itself, and its connections that store the information and not the individual nodes. Neural networks are built to adapt to new information, or for that matter to the change of configuration of nodes. If we work from a model of human knowledge and learning as being based on neural networks, then any organization is a network of neural networks, an internet, and we all know that an internet is really just another network. So, we can see an organization having knowledge that the individuals don't have. To make this a little more concrete, take a look at how RAID 5 disk systems work. The information is spread across several different drives with some redundancy built in. No single drive contains a complete set of information. It is more efficient (depending on the type of application) than simple mirroring. It is also more secure, in that if one drive is stolen, it is extremely difficult to ascertain the data stored. This would carry back nicely to Mark's original parallel between a terrorist cell and a distributed network. A well set up organization should have information redundantly distributed, but in such a way that no single node can be used to ascertain the information of the organization. I must admit, however, I don't have any experience with terrorist cells, so I don't know how close they come to such an approach. As to the despoiling of natural resources, it is my recollection that this was explored quite nicely in a science fiction story a few decades ago, where they explored the social implications of teleportation. I believe the story was named something like 'Flash Crowd'. Does anyone remember the story? Aldon __________________________________________________ Terrorist Attacks on U.S. - How can you help? Donate cash, emergency relief information http://dailynews.yahoo.com/fc/US/Emergency_Information/
What struck me most at Maneul Castell's talk at AOIR last year was his emphasis on examctly the opposite of this "distributed networking" happening with respect to business locations. Cities like SF and NYC, and regions like Silicon Valley, were growing denser and denser, just as the technologies they were developing seemed to promise to obliterate distance in time and space. As I remember, Castells concluded that there was some fundamental level of "human capital" that was necessary for the face-to-face engagement that's still at the core of even postmodern business deals. So, the more spread out the world gets, the more centalized these "nodes" must be, to serve the periphery. At the time, it seemed ironic, almost nostalgiac; now, it does seem doomed. I imagine corporate headquarters becoming the same kinds of paramilitary "gated communities" described in Davis's City of Quartz and Neal Stephenson's Snowcrash. Of course, Castells also hyped the wonders of Cisco's business model, just before their stock tanked. Any other thoughts on how the attacks may change the geography of America? -- Ted Friedman, Ph.D. Assistant Professor, Department of Communication Georgia State University (404) 463-9522 tedf@gsu.edu; ted@tedfriedman.com http://www.tedfriedman.com
At 03:20 PM 9/17/2001 -0400, you wrote:
Has anyone else on this list been struck by the parallel between the current portrait of de-centralized terrorist networks emerging in the media coverage and the organizational logic of distributed computer networks? [...]
"It takes a tank to fight a tank. It takes a network to fight a network," says John Arquilla, senior consultant to the international security group Rand and co-author of the forthcoming "Networks and Netwars: The Future of Terror, Crime and Militancy." I would submit that if notion of networks is a key elements of Internet research and theory that is also one that needs much further theoretical consideration. Those that have responded to the note the initial note about networks I think are in agreement. I cite Arquilla in part to indicate that we need to consider networks as a more complex entity than as a transport and storage system of information and connectivity, such as to what I have sometimes called flat or planar maps of mere connectivity. Can we take social network theory as a preliminary phase in this inquiry, and look to an emerging network theory, indeed to a theory of both network emergence and network control? Can we deal with the networks interacting along a variety of scales, time frames, and evolutionary possibilities? It has been only with great difficulty that many theories of media have grappled with the new network based media. My own experience was such that the 'broadcast model' simply focused on the power of corporate media to deliver content, and the parameters of personal and inter-personal behavior and semiotics involved in 'receiving' it. Hence, when network based media became ever more critical to media studies in the late 1980s, there was a lopsided battle between those who would position new media as bearers of democracy (cf. Benjamin followers) or fascism and corporate control (cf. Adorno followers). Mark Poster has had some interesting things to day about this 'stalled dialectic.' My research suggests that we need to put into questions the very definition of networks. The definition of the network, its role, and activities, is up for grabs, and is more complicated than it initially seems. The question of terror networks brings this realization out into the public- it is an issue not easily addressed by the traditional categorical thinking that leads to the identifications associated with declarations of war (or research). The studies of the telephone and of social networks is important, but I want to emphasize that we must include issues of epistemology/semiotics, 'multi-dimensional dynamics,' and questions of power of and in networks to the mix. I really don't know where this sort of deconstruction of the notion of the network is taking place- but given that I am preparing a lot of material related to this (indeed my diss. took up this theme), I am sure that there are many others out there as well. While the Internet research community and network researchers in general have taken up and have begun to explore the theoretical and practical problems of locating identities, agency, resources, bypass strategies, etc. in an increasingly networked world, there now seems a need to present some of these findings in a way that can inform public policy. It is as if many of the leaders were bound to outmoded conceptual frameworks in calling for this 'war' whether we look to their activities from a somewhat cynical, opportunistic perspective, or from a more moral, ethically 'pure' perspective. In my own theoretical work, I have looked at questions of 'nested' identities- what happens if 'parts' don't exactly match up to the 'whole' - what happens if individuals can't be located by a single nation or category- because they have become more connected to an information or transportation network: War- but against whom? And if we are to think of these networks in terms of internally coherent, even autopoietic systems, then how are such 'networks' bounded, and what keeps one individual as a member of one such group, rather than another. I look not only to endogenous factors, but also to external factors, and it is the play of the internal to external factors that needs a lot more attention. So where is the end, or limit of a terrorist network? Suddenly this become an important, legal question- and I can imagine the trials of some people who gave money say to a religious institution arguing that they do not believe or agree with the 'agency' attributed to the network- are all the members of the US linked to US government actions- who to the activities of networks become characterized as unified as if that network were a single body. Isn't the obvious link to these questions and those of semiotics and categories also interesting? There needs to be promulgated new ways to think about networks. The issue as to whether networks concentrate or disperse people and resources is an old one. I would point not to Castells or Saskia Sassen but to Ithiel de Sola Pools' classic 'Predicting the Telephone.' We find that telephones were simultaneous predicted to cause both concentration and dispersion, and in fact they did both. Combined with a transport technology (the elevator), we find that telephone's promoted concentrations of works in multistory buildings. With the threat on them as obvious targets to terror, they may become more dispersed. But companies need 'pools' of labor, and there are economies of scale, supply issues, and security/privacy issues associated with having companies in close proximity, in association with other dense conclaves. The idea that one responds to the threat of a network with a network demands consideration. The netwar strategist argue that given the need for 'trust' within networks, that is pays to in some way to infect the network with distrust. "It's a whole new playing field. You're not attacking a nation, but a network," says Karen Stephenson, who according to author Joel Garreau studies everything from corporations to the U.S. Navy as if they were tribes. It is interesting to see how redundancy and decentralization in the financial networks appears to have limited the loss of one 'node' - but at the same time, we should be as willing to ask, just what was unique about this node and its environment, and how will its transformation in turn transform the larger systems to which it is connected, and which it helped organize. We have to move beyond the Leibnitz world of monads- nodes are not simply passive, and they also can exert a organizational influence over the whole. Just as we can look to the ability of nodes to relay information redundantly (as do libraries in a fashion), we need to consider the uniqueness of the node, and the complex cybernetic environment in which any node both works and is empowered. These questions will have strategic importance as we move to new paradigms by which to consider war and media. Willard (mailto:willard@well.com)
Certainly both forms of networking emerged with a similar goal in mind: a resiliance to the forms of centralized large-scale attacks characteristic of warfare in the first half of the 20th century. Redundancy and de-centralization are defensive structures whose effectiveness is demonstrated by the fact (reported yesterday, I think) that despite the destruction of something like 10 percent of Manhattan's office space, suprisingly little data was lost. In the face of this kind of distributed networking, the type of military response envisioned by Bush/Cheney et alia seems disturbingly out of joint. Just as we wouldn't imagine that we could take down a network by hitting a node, so too does the goal of "taking out" Osama bin Laden seem more symbolic than effective (to the effect that it works to proliferate cells of resistance, it might even be read as counter-productive). I'm wondering if there's some way to use the commonly accepted discourse on computer networking to shed some light on the current debate over the appropriate U.S. response to the recent acts of terrorism.
participants (8)
-
Aldon Hynes -
Christian Nelson -
Nancy Baym -
Prof. Andrejevic -
Steve Jones -
Steve Jones -
Ted Friedman -
Willard Uncapher