A. Introduction: What's true for e. coli...

Primitive operators 'scale' and 'convention' are introduced — Mice and Elephants compared in the context of scale, and related to discussions of the economy, the internet and Santa Fe — Scale becomes intransitive verb — Convention explained with respect to community, communication, empire, and telecommunications — the internet is identified as an Important Difference within convention — Convention and Scale form the beginning of a beautiful relationship — the location of the internet is sought, and not found — Various Philosophical Musings.

La conventientia est une ressemblance liée à l'espace dans la forme du "proche en proche." Elle est de l'ordre de la conjunction et de l'ajustement. C'est pourquois elle appartient moins aux choses elles-mêmes qu'au monde dans lequel elles se trouvent. Le monde, c'est la "convenance" universelle des choses.

Conventienta is a resemblance connected with space in the form of a graduated scale of proximity. It is of the same order as conjunction and adjustment. This is why it pertains less to the things themselves than to the world in which they exist. The world is simply the universal 'convenience' of things.

Michel Foucault from "La Prose du Monde" in Les Mots et Les Choses

1. In January of 1999, at dinner with Sean Doyle and some friends, Sean told me about an article from the New York Times 'Science Times'. The piece [Johnson99] described mathematical models of quarter-power scaling for biological entities, like mice and elephants. As organisms ascend in absolute size, their relative metabolic rates scales as the 3/4 power of that. The mouse's heart rate beats the mouse's size far faster than does the elephant's. If you posit a spherical mouse (Sean said: "You have to make some simplifying assumptions"): surface area grows in two dimensions, but volume in three. Mice and elephants' ratios differ (surface to volume), offering a possible explanation for their heart rates and metabolisms [1]. As in so many recent cases, we can thank the willfully wacky folks in Santa Fe for insisting on research into what was obvious to many to find the non-obvious answer— that self-similar fractal networks of vessels of invariant size (such as the smallest capillaries) might have something to do with blood pressure, heart rate and metabolism. So bigger is better, it turns out. Or at least, bigger is better if you are a dinosaur fan: for Sean, this indicated that dinosaurs were actually very efficient creatures, and probably very long lived. The story was one of hundreds that Sean and I have traded over the year or two that I have been watching Amicas, but by some peculiar combination of math and animals, this story stuck.

2. Such narratives birth powerful analogies. Small strong ants (start-up companies) can lift hundreds of times their body weight, while lumbering giant octopi (multinational corporations) have no such strength, but live for decades and routinely crush smaller animals on their route. The metabolism of small companies is high, burn-out is common, but they extract maximum energy from minimum input (Amicas employees sometimes go months at a time without paychecks, foraging in the wild for credit), the metabolism of large companies is slow, a limit case perhaps being the University, in perpetual hibernation, or at least molting. But in the case of Amicas (or internet start-ups specifically) the image is more than analogy. Not only are convenient metaphors good business plan material, but in fact the substance of their work: the particular problem of the technical (structural, organizational, and communicational) scalability of institutions, organizations and 'enterprises,' and their speed and stability as they grow is less analogous than homologous. Something tricky about the problem, and indeed, a certain class of scientific investigation conducted mostly in Santa Fe, New Mexico, seem to bring together the biological and the computional problems [2] into an object lesson that challenges the creative imaginations of entrepreneurs.

3. In recapitulation, the analogy-cum-theory quickly extends outside of the firm to the market. We recognize that firms and organization grow. From the Sherman Anti-Trust act on, the twentieth century has been one long experiment in pure size, corporations representing the pure products of American bigness [3]. Markets scale, however, and in the modernism of the corporate era, the size of a corporation allowed it to reach "economies of scale." Today, such a relation leans more on complexity theory than brute growth. Bigger is not necessarily better, diminishing returns to scale— that gravitational constant of industrialism that was Marshall's natural constraint on size, has given way to "increasing returns" to scale and the creation of "economies of scale" from the smallest capillary units. Absolute size and reach are no longer related. Markets are made, scaled to order. Scale is an intransitive verb. [4]

4. The technically specific component of this change is the internet. The internet scales also. Internet servers are scalable when the technical constraints on one transaction are the same as those for a billion. No need for a billion servers to serve a billion hamburgers, this baby scales. The intersection of a certain class of problems represented by the Santa Fe Institute with the history of computing would demand a more complete genealogy [5]. The models, metaphors, and mixed messages that go by the sundry names open systems, distributed processing, distributed computing, enterprise computing, distibuted object-oriented computing, object management, complex systems (the list scales) is part of a more general scientific and technical interest in the problem of large, complex organizations— and at one end— in theories of self-organizing behavior of which the market and evolution are the two most familiar and commonly repeated exemplars.

5. But scale is only the login to the present. Convention is the mot d'ordre [6]. Convention will be the password to a single file of terms that lie, appropriately, in convenientia, step by step: agreement, standard, contract, custom, convention, regulation, law. At various points, this section touches on all of these terms, in specific contexts where they appear, and as ciphers for a more general theory of the social bond. In particular three related "negentropic eddys" (to employ the fanciful language of "complexity theory") form around convention.

6. One. Community was once the scale of convention— the space of sharing. Community is an absolutely devalued word today, it is agreed, indicating palpably only what is gone [7]. Community is an emblem of scale as well, in particular the small scale opposed to corporate bigness— the space of authenticity and presence where a moral frame orders obligations and responsibilities by virtue of conventientia— propinquity. The "scientific community" in particular is the example that appears and appears here. What is it that allows sharing in science— sharing data, sharing techniques, sharing ideals, sharing knowledge? For Weber it is a vocation, a calling that comes from the severity of desire. For Lyotard it is sharing based on dissensus, a calling to outdo in an agonistic game with mutable rules. For many scientists it is the dream of a perfect language, a final theory, or the revelation of secrets— callings no less powerful for their mythicalness. There is also sometimes a desire to share what is learned or discovered, to transmit, to teach. The discovery or solution of a problem is only half the desire, because it is the explanation and teaching that actualizes it. Once the university was the locus of this activity, but in the very recent past, the internet has taught a new generation of scientists to come and go with vastly different speed and scale. The community, such as it is, is stable by reference only to a set of shared interests, unsolved problems, articulated concerns, not by reference to people or place. The usenet group, the ftp site and the free-software model of software development (the Linux and GNU operating system) are the models of collaboration for a genearation that is migrating "science" outside of universities [8]. The boundaries of the university are denaturalized, rexognized as the arbitrary barriers to solution that they often are.

7. Two. The convention of communication that the internet represents is not new. It is, in Bernhard Siegert's terms, part of "the epoch of the postal system." The engineering of empire has always included systems of control and communication whose model is the post. "Reason must be delivered to consciousness." [9] Its successes and failures in the past have depended on cooperation and agonism, assent and persuasion. Communication itself [10] is the metric of freedom for the subject whose face, in the sand at the edge of the sea, has washed away to reveal grains of sand as silicon wafers, or as bits. To communicate freely? But what does this mean? Every example of a communications 'network'— roads and routes and posts and messengers and secrets, every example of an empire's scale from the runner at Marathon to the cipher of Enigma to the Global Positioning Satellite— is recapitulated in the internet. It's what makes a thousand just-so stories of its origin make sense to a thousand new users— the internet is TV, the internet is newspaper, the internet is nuclear-war proof military network, the internet is post office, the internet is hypertext (secret message) and cyberspace (outer space). There is an historical imagination here, or an imagination of history and the world as totality. The keywords of this epoch were development, growth, progress, evolution, revolution, modernity. This was conventional modernity.

8. Three. The convention of the internet is different. The standardization of the internet, the particular technical standardiztion that allows for its scalability is in fact the only thing that has the dubious distinction of being called 'new'. The method of standardizing is a pragmatic and fluidly programmable one. It is strictly, explicitly opposed to standardization by empire. Or perhaps it would be safer to say that the internet is empire by other means. The moral equivalent of imperialism. These standards (e.g. TCP/IP, HTTP) have achieved world-wide agreement through a process that is less explicit than 'evolutionary' (though if it is, it is Lamarkian). Not only has it leveraged itself onto an existing network of institutional connections (e.g. universities and labs first, then corporations), but in its modiafiability, threatens to change them each time it changes. This pragmatic, modifiable method of 'agreement' has riven 'us'. The internet— this new space— this public sphere that is outside of any nation, but 'in' all of them, that is outside of any corporation, factory, university, institution, home, but 'in' all of them, highlights the connection between location, value, and the first person plural of organization in all forms. 'We' are no longer the same 'we.' The 'we' that decides on agreement is no longer primarily spatio-temporally together, it is an asynchronously transnational 'we'.

9. The conjugation of scale and convention is intended to act out the connections between culture and locale, where the internet, and its standards in particular, serve to signal the end of the conventionality of such connections. The empirical question is always located. The location of the firm, the location of the hospital, the location of work, the location of suffering. No methodology without constraints, no ethnography without place. As the world turns, the young and the restless are at home and away everywhere. Migration, casualization, increased tourism, Rupert Murdoch, all these simple things militate against culture being bound to place. Culture is packed up, moved, installed in new places, its tenaciousness is well known— it inheres in performance, in repetition, in formalization. It resists, it transforms, it syncretizes, it gives meaning. Sean Doyle's joke about standards somehow depends on this notion of culture as diversity: "The great thing about standards is that there are so many to choose from." However, in the worlds of telemedicine, internet healthcare, and corporate execution, culture is absent (and certainly not only in the healthcare industry). It is either 'in museums' or it is banalized into a management strategy ('corporate culture' or multiculturalism). Standards are generally experienced as the furthest possible thing from 'culture'. They are technical problems solved by engineers.

10. But what if there is a conceptual continuum between technical standards on the one hand and 'culture' on the other? Then technical standards are more than just technical solutions to arcane problems, and more than simply agreements achieved by various means (however conventionally political or customarily scientific). And 'culture' therefore demands a description adequate to the technical components that constitute it— whether religious conventions, or international organizations for standardization— a description that is not a scientistic reduction to formal problems of coordination and control (systems theory, game theory, convention theory), but which resists the metaphysical implications of its indiscriminate application (the hermeneutic temptation to attribute the ephemeral ineffable to 'culture' or 'meaning', to close the circle by reference a center outside itself [11]). My stab at this comes via the internet in its scale (world-wideness) and convention (standards).

Where is the internet?

11. One way to answer this question is the way Carl Malamud [Malamud92] or Neal Stephenson [Stephenson96] ask it: as Hacker tourists who travel the world sampling the twin delectations of terrestrial cables and exotic cultures. Anthropology's monsters, these two in their respective accounts provide readable, albeit insufferably self-aggrandizing portraits of the most material aspects of the internet. These cables, routers and hubs do not quite connect to the keyboards of hackers with less wanderlust and more prosaic pleasures. Here the location of the internet is a technical question, much like the location of empire. Malamud's book traces the outlines of the question of standards raised here. His mission was to take the standards published by the International Telecommunications Union— standards needed by engineers all over the world— and post them on the internet. In 1992, this turned out to be a Sisyphian task, and Malamud's impatience makes the story of his globtrotting techno-ethnography of early internet homesteaders a classic example of Hacker Humor, a mix of obnoxious and perceptive, ultimately tending towards self-promotion. It is a story of the misfit between the open ethos of information sharing of the internet and the closed hierarchical, decision-tree oriented bureacracy of the ITU. It will reappear further down.

12. Stephenson's travels have a richness that wavers between technophile novelist and wannabe hacker, mixing the obnoxiously perceptive with the historically portentious and the fictionally impossible. It is a journalistic corrective to the much more sober catalog by Daniel Headrick, The Invisible Weapon[Headrick91], which details the reader to death with the pre-history of the international web of wires and their role in the threat and cracking of secrecy. Stephenson's stories rely on a sharp cut between real and cyber space, or as he prefers referring to it: 'meatspace' and 'cyberspace'. His method is to follow the laying of roughly 28K among hundreds of thousands of kilometers of cable laid in the 90's. In particular, the FLAG cable laid cooperatively by AT&T US and KDD in Japan. Again, this kind of locating, although essential to a demystification of cyberspace as some total cinema of the mind, mystifies in another way by ignoring the non-glass, but no less material software (and standards) that runs both the laying of the cable and its subsequent uses. If Kittler can insist that there is no more software, does this not imply its opposite as well [12]?

13. Another way of locating the internet might be the psychosocial mysticism of current scholarship on "cyberspace" ranging from sociologists and psychologists studying the dynamics of interaction in MOOs and MUDs, where avatars are imagos for a new era of psychoanalysis to the strange connections of 'hypertext scholars' such as George Landow and Jay Bolter who insist that cyberspace is little more than a literalization of 20th century French philosophy [13]. For these studies the internet's location is somewhere in between the screen and the mind, in which self-contained egos learn pretense in the imaginary space of role-playing that cannot ultimately be materially differentiated from novels or theatre. These treatments do not yet ask how we might share the internet, much less how we already share "culture." It perhaps unfair to demand of this scholarship an answer to the question, where is the internet? However, in some cases this scholarship will call itself a "sociology of the internet," (begging the question of where 'society' is), and in others it engages a long-standing philosophical discourse on reality and appearance, or on the platonic idea and its location- yet for all that it rarely references 'cyberspace' as a cave, much less as a physical criss-cross of copper, glass and waves, layered with software and bits.

14. A third way of locating the internet, and this is the most promising and least attended to, would be to chart the political economy of manufacturing that is the absolute sine qua non of the internet, whether it be its cables, its software, or most importantly, its semi-conductors. The clean rooms and home-work, the international division of labor and the forgotten interstices of a development consensus that is neith in nor yet on the internet, but nonetheless controlled by it [14].

15. This corporate map of the globe is in direct conflict with the globe that the internet will become, especially in the hands of a vulgar version of development economics that would demand that economies all over the world move through the various stages of capitalism in order to achieve modernity and the promised land of information society. The conflict precipitates non-synchronous nows all over the world, regimes of power and decision-making focused on a past imagined in the present, committed to the impossibility of the future that organizes these regimes. Modernity and revolution locate themselves in states, or at least in societies, whether those associations are professional or class based, collectivity dominates the promise of progress. But the internet's focus on singularity, its game-theoretic model of collectivity is strictly opposed to progress of this type. History, or its end, becomes one preference among others, awaiting revelation.

16. Materialism demands that the internet reveal its innards. Theology, its spirit. But can the internet be a thing, or not be a thing? Like The Thing [15], it is from another world, yet it gets inside you. Once there, the thing takes over, its only mission to contract with another you and make more things. Is The Thing society? Joe Dumit directs this thought by a reversal: not the social life of things but the thingly life of the social. Not just the commodity as fetish (society in misrecognition) but society as the unfetishized, the not-yet thinglike but living substance (or even, that which fetishizes humans) [16]. Stability exists no where in society, least of all in what is blithely imagined to be 'the medium' (not always the message, sometimes only the medium, pace Freud). The internet is society, but it is a society that a recognizable 'we' no longer share in the same way. The internet is also culture, as I hope to show with respect to convention, that is standards and their role in cooperation, contract, and conflict; but more even than that, the internet is the world. What 'we' once invesitigated as limits of the world— life-world, experience, the phenomenological reduction that craved time and space as its factors and its ends— this is no longer 'our' world (and the relationship between the first person plural and the world as image of totality is what I can only offer as conundrum). I risk a constant mystification here, either by reference to a philosophical idiom that treats itself to universality via this notion of world (or end of world) or by a more dangerous reference to the instantly devalued heraldings of digerati-scholar-journalists who make promises of an unimaginable utopia for an 'us' that patently excludes. All of the ways in which the "world"— from world-views to Weltbilder, to the wide world of sports, to the endless dissemination of the "global"— are used to signify some kind of New World are narratives of a kind of spiritual, or theological, progress which is strictly incompatible with a re-description of material aspects of the internet and its users (in this, I think the philosophical idiom is not far removed from that of the digerati, both yoked to an obscure imperative that cannot risk stating its own desires). Yet, the internet is not just a thing to be mapped or enumerated (it does this itself, with vast speed and considerably more detail than the best researcher could achieve), but a problem of substance and control, like society, like culture. As a result, perhaps even the very question "where is the internet" (like 'where is culture?') has become impossible to ask, much less to answer.

Last Modified 11-Sep-99 9:01 PM ckelty@mit.edu

Go Back to the Start