On any other day I might even accuse myself of quibbling. However, tonight I’m going to insist, in the spirit of serendipity, that the distinction between the two terms as used is broadly similar to the contrast that Eco makes between an ordered list of things and the diverse, unique, incommensurable things-in-themselves that are represented therein. In this instance however, it’s the list-like consistency and cartographic order implicit in “geography” that’s the source of (false) reassurance. Arguably, the more narrowly gauged “terrain” conveys the same sense of place, albeit without any implication that (this) place is similar to or reachable to/from any other — which seems like an apt description of the Internet… at least from where I sit ;-)

TV

Tom, you’re certainly right that if it’s a geography, it’s a totally weird one. But then the question is why — despite the dissimilarities — our vocabulary for it uses geographic terms so frequently. There seems to be _something_ about it that is geographical. For me, I think it matches the elements of geography that have to do with the meaningfulness of place, not their arrangement in space. Maybe.

One other thought for now. Even if the institution of copyright had never existed, and/or the Internet conveyed *nothing* but real-time/synchronous communications between mutually consenting individuals, the Internet would still represent a major challenge to pre-Internet ideas about privacy and autonomy. Structurally, the protocol-defined and standard-ized features that make the Internet work also make it vulnerable both to different kinds systemic imbalances, as well as to user-initiated manipulation and abuse. In this sense (and many others, actually), Internet protocols are exactly like the monetary instruments, financial flows, and institutional features that are supposed to enhance and sustain liquidity in the conventional economy. Want to have a currency that is 100% proof against inflation and deflation, and that will never be vulnerable to theft, debasement, or counterfeiting? The solution is simple: just adopt as currency something that nobody else will ever recognize or accept in an exchange transaction. Alternately, if you want a currency that you can use — i.e., that actually works as a liquidity mechanism — then the best you can do is mitigate (perhaps very substantially) those systemic and specific / misfeasance and malfeasance-related risks. This is equally true of Internet protocols, for exactly the same reasons.

However, what does make the Internet different from the world of conventional exchanges is precisely its lack of any fixed “geography.” Intentional bad behavior and innocent but careless mistake made by one party can have the potential to significantly impact individuals and activities regards of the intervening distance between them. Clearly such risks are not absolute, but in a very real sense attachment to the Internet conveys the power to create potentially large effects “at a distance,” quite often regardless of the knowledge or consent of those effected, quite often in ways that cannot be naturally or directly attributed to any efficient or final cause (i.e., the kind that might indicate the source of the problem).

Without at least the technical possibility of accountability, such power would not be sustainable for long — and some level of “identifiability” is a necessary prerequisite for that possibility to exist. Because of this, perfect anonymity can never be anything more than a temporary, domain-specific possibility, not in the current Internet anyway.

While that might make some strong privacy advocates cringe, I suspect that the only other alternative — an immutable, strict construction Internet that resists all changes unless that have not been fully specified and approved in advance — would be far worse, assuming it would be viable at all.

TV

Privacy issues notwithstanding, the Internet can only be viewed as a “geography” if one is willing to accept that that term can have a secondary denotation that is basically antonymous with its primary definition — i.e., that it can refer either to a relatively fixed, consistently and reasonably well-ordered arrangement of features**, or alternately to an immensely plastic, self (re) ordering collection of features** that is generally insusceptible to any consistent, independent ordering scheme over any extended duration.

[[**"Features" in this case meaning physical or logical "pipes" as well as the things they convey and to which they are attached]]

That said, some elements of the post-structuralist worldview actually have real-world technical relevance/applicability in this domain, at least in the narrow sense that every perspective and empirical measurement of the Internet is (literally) a view “out” from a particular vantage point onto a distinctive terrain (a.k.a. loosely bounded set of features**), and every such perspective is only partially commensurable (at best) with the equivalent view(s) from anywhere/everywhere else. This holds regardless whether the perspectives being compared are simultaneous or are separated in time.

Interesting talk — TV