this might be of interest - older piece - Alan
Alan Sondheim
Linguistic Dispersion and the Theorization of the Internet
While the Internet itself is a collocation of client-server technologies
that are not necessarily implemented with the same structures (even TCP/IP
(1) is of limited presence), there are elements that are open to analysis,
that cross traditional boundaries, and that are susceptible in fact to a
theoretical linguistics of language change, acculturation, and upheaval.
Colin Renfrew describes a number of scenarios for such transformations,
some of which go back to nineteenth-century thinking. In any case, these
and others can be used for theorizing elements of the Net that otherwise go
undetected or are described purely through phenomenological approaches.
[See Renfrew, 1987; Hodder, 1990.]
I will give examples of these scenarios and their usefulness in theorizing
the Net. I am not of course vouching for their verity in Indo-European or
other studies, although my understanding of them has come from such read-
ing.
1. Takeover: One domain is subjected to another by conquest, and the latter
may impose new standards. An example is the Netscape takeover of Mosaic
(and for that matter, Netcruiser); early Web accounts presented the latter
as the primary browser. (2) The takeover is accompanied by continued
development of Netscape (see below), into further extensions of html, Java,
and now telephony. (3) This can be considered a _replacement model_ in a
network _core._
Takeover results in _colonization_ that then transforms into
self-colonization, a process described by Franz Fanon. [See Fanon, The
Wretched of the Earth.] In self-colonization, also referred to as the myth
of a national literature under the sign of empire, the colonized introject
the values of the colonizer. Thus Mosaic has had to make concessions to
the continued growth of Netscape and its dominance; and thus the various
commercial BBS-services (4) such as Compuserve, AOL, Prodigy, etc. have
had to scramble to "catch up."
Takeover on IRC, to give another example, can refer to channel takeover,
(5) and if new techniques are developed, they're recoded and implemented
in the various war programs.
2. Language diffusion through perceived superior technology imported from
outside. A Hittite document on horse-training uses Mitannian terms, for
example. [See Gurney, 1990.] Basic Internet terms are often borrowed by
other languages directly, just as Quebec _joual_ spelled "weekend"
"ouicenne" in some key texts of the 60s.
The import of technology occurs in a number of areas - from Java applets
imported into home terminals, to the IRC war programs mentioned in #1. A
strong instance of this is the continuous cross-fertilization of MOOs, with
programming on one MOO ported to another. (6)
Another example is shareware/freeware itself, which is a literal import;
tiny fugue goes from a single program to "tiny-fugue-like" clients. It
becomes a class term - just as, for some people "listserv" refers to the
totality of email lists, not just those employing listserv software. (7)
(Language diffusion can be extremely complex. Diffusion can be localized
and in relation to the Net may include other communications modalities
such as the telephone, printed matter, snailmail, and face to face
contact. Think of hacking circuits which use all sorts of contact,
including hacking conferences, 2600 magazine and 2600 newsgroups, etc.
These modalities can be seen in relation to Fidonet, (8) as well as local
bulletin-board software, databanks, chat exchange, and in-house email. All
of this is describable by the "wave of advance" model described by Renfrew
in Archaeology and Language. Random localized movement (for example,
farmers' children settling near their parents) results, through quickly
increasing population density, in widespread cultural influence - without
migration, takeover, etc. This model works well with the modalities
mentioned above, all of which could be considered operative on a
"neighborhood" model.)
3. There are also _autonomous_ or internal transformations which always
occur in a series of steps or _plateaus._ Beta-versions of software are
the simplest example, or decimal-incremental indexing of continuous
up-grading. (I'm writing this in Pico 2.5 for example.) (9) Beta-versions
often results in major changes, and the integers are reserved for basic
major changes (Netscape 1.0, 2.0, etc.). This numbering does go back, I
believe, to Wittgenstein's Tractatus and the Principia Mathematica of
Russell and Whitehead; there may be earlier antecedents.
Plateaus "slough" off into _minor steps,_ which are often created to fix
software bugs. These steps may be nothing more than patches into the
software, or they may be integrated into a new beta version. Usually, they
are released before the new version, however, sometimes with a warning.
They can be an Internet form of recall; I've seen this from Pavel Curtis
in regard to the MOO core (10), for example. I consider the minor steps a
form of _accretion,_ the slow growth and maturing of a beta version before
its abandonment for the next plateau.
4. Renfrew describes a process of empire collapse which is critical for
understanding the Net. The boundary regions of an empire are always sites
of contestation; they also represent the phenomenological horizon of the
Other at odds with the core, which can be conceived of as "the civil." An
empire may over-extend, however, in which case, the boundary regions apply
pressure to the core; at the same time, they may begin defining themselves
as autonomous, using the tools and languages of the core. The core begins
to retreat from the peripheral zone, and the boundary regions move in to a
limited extent, becoming independent and often anarchic. This process
(which I have elaborated on above) characterizes to some extent the
break-downs of the Roman empire and the USSR; in both cases, anarchy has
meant a problematization of etiquette and rhetoric (both interrelated) as
well as language itself. (I use "anarchy" in the sense of confused local
rule, often chieftain-organized, and often in a state of competition with
other local groups.)
If we look at the Net pre-1989, for the most part it is small, organized
(the RFCs (11) playing a major and critical role), and, in spite of the
Worm of 1988, (12) relatively pacific. The development of the Web took
this Net by surprise; the Web can be considered an intrusion or migration
of new standards (html, .gif (13), etc.) across the older Net (within which
standards such as .gif were relatively well-defined in limited domains).
What happens?
a. The Net as a whole splits into ASCII and Web-dominated applications,
with other sound/video/text applications on the periphery. (14)
b. This split parallels the split between community and presentation - as
well as the split between information and entertainment, and the split
between active and passive. (Please note this is _extremely rough here,_
and not to be taken at all as "ultimate" or "fixed" characteristics. The
reality is much more complex.)
c. The older Net becomes less coherent, and more anarchic; there is a
general system collapse as advertising, hackers, spams (15), viruses, etc.
invade. RFCs no longer have the governance (which was by consensus only)
that they once had; their non-commercialist aspect is bulldozed over.
Newsgroups and the like are increasingly invaded.
d. In response, this older Net, what I call the darknet, develops more
sophisticated responses, such as filtering, war defenses in IRC, and
totally moderated newsgroups (16). These parallel the development of
monasteries in Europe, "citadels" of learning preserving classical texts
and even literacy during the "Middle Ages." As with the monasteries, one
can say that the moderated newsgroups, etc. are interrelated, but the
general sense of Internet community has been increasingly breaking down.
In other words: We can see the older Net collapsing with the rise of the
Web. The collapse is accompanied by a lack of central authority, and the
rise in anarchic tendencies, coupled with defensive responses. And instead
of seeing this as a fight between anarchy (in the traditional sense) and
capital (which it is), one can also see this as a transformation of
language itself - the older languages becoming jargon, and the newer
languages adopting other sensory modalities, such as image, sound, and
movement. (It's useful to think of "jargon" in a technical sense here,
referring to a somewhat underground/underrepresented demographics which has
created its own speech, built on archaic forms to some extent. Consider the
jargon on IRC, or MOO/MUD talk in which the prefix "@" can be used to
delineate an ingroup, create a sense of "local color," and refer to
technical issues. But this jargon, like the MOO language or the MUD LPC,
is becoming increasingly "minor," in the sense of a "minor literature"
(see Deleuze and Guattari on Kafka) - all of these are _regional issues,_
far away from the concerted sense of power and sentience displayed on the
Web at this point in time. (17))
This analysis is crude as it stands, and the processes outlined above are
never "pure." But the analysis does take into account such things as
protocol totalities (Unix, html, TCP/IP) and their borders, the growth of
new applications and the disappearance of older ones, and the importing and
exporting of code across the Net, even across applications. If the Net is a
widely disparate accumulation of interconnected functions, protocols,
applications, and demographics (not to mention telecommunications
channels), it is also a somewhat closed region (in the sense of a closed
set) of competing languages, functions, protocols, etc., with _traderoutes_
of shareware/freeware/coding among them. Models of linguistic dispersion
can help one understand this relative cacophony. The modeling is rough, but
it just might be the best we have.
(Further research would take into account "core borrowings," basic
structures or functions that would be imported and rewritten from one to
another application. The various _forms_ of applications or functions
would be considered, much as differing literatures are examined. Again,
Laurel's work on Computers as Theater [Laurel, 1993.] has some relevance
here. Finally, there is the issue of dialect. Chomsky somewhere describes
a language as a dialect with an army; here, we can consider its equivalent
as an application with corporate backing. In any case, should Netscape be
considered a browser language, Mosaic and Chameleon as dialects? While
these terms seem absurd, tracing influences among the applications is not,
and here again we would turn to linguistic dispersion theories,
considering not only older tree models of development (i.e. Netscape 2.1b
coming before Netscape 2.2b etc.) but also models based on diffusion, wave
phenomena, and overlapping regions.)
==================================================
Footnotes:
(1) TCP/IP is a _protocol suite_ or software bundle that defines most of
the Internet communications frameworks.
(2) The Web is the World Wide Web, a hypertext formatting for the Internet
which is generally accessed through a _browser,_ a software program that
can read html (hypertext markup language) documents that are on the Net.
Mosaic and Netscape are two graphics browsers; Netcruiser is a graphics
browser proprietary to Netcom, an Internet service provider.
(3) Java is one of a number of new languages appearing on the Internet,
used in conjunction with the World Wide Web. It allows for programming
(based roughly on the C++ language) everything from animations to word-
processing programs.
(4) BBS-service: Bulletin-Board service that provides front-end proprie-
tary materials, as well as Internet access.
(5) IRC is Internet Relay Chat, "live" chatting with a number of people at
once. Users talk on channels; hackers can "flood" them by generating huge
amounts of useless data at high speed. War programs are used for automated
flooding.
(6) Applets: Small programs that are used in conjunction with Java and a
home computer; they are formatted and used only when needed. A MOO is a
MUD (Multi-User-Dungeon)-Object-Oriented, a text-based virtual reality or
world, within which participants may program rooms and objects of all
sorts, talk to one another, and even send MOO email back and forth. There
are all sorts of data-bases within a MOO, including internal email lists
and libraries of text documents. (A MUD is based on role-playing adven-
ture games, such as Dungeons and Dragons, and a MOO is a social extension
of a MUD.)
(7) Freeware: Software programs available for free on the Internet.
Shareware are programs which come with a nominal charge attached. Tiny
fugue is a client or "window" through which one accesses MOOs or MUDs.
Listserv is one of the most popular email list software programs.
(8) 2600, The Hacker Quarterly, is just that. There are 2600 newsgroups
associated either formally or informally with the magazine. 2600, by the
way, is the frequency of the standard telephone signal. Fidonet is a BBS
alternative to the Internet which has a strong following even today.
(9) Beta version: Text version of a software program, released to the
public for evaluation. Pico is a simple text editor developed for the
Unix operating system.
(10) MOO core: The "heart" of the MOO software, which runs the applica-
tion. There is also the MOO database, which contains the specifics of the
MOO world itself, as well as the list of players, passwords, and so forth.
(11) RFC: Requests for Comments - a group of indexed papers defining and
discus- ing the protocols used for Internet communications, etc.
(12) Internet Worm: A program which "escaped" onto the Internet; like a
computer virus, it spread, slowing down traffic until it was caught and
defused. The incident pointed out the relatively lax security of the Net
in the late 1980s.
(13) Both .gif and .jpeg are formats used for sending image files across
the Internet; Netscape and other browsers translate the files into pic-
tures.
(14) ASCII is the standard for text-based Internet applications. The
division here is between "purely" textual and extended (audio, video,
sound) applications.
(15) To "spam" an email list or newsgroup on the Net is to either flood it
with a huge quantity of useless posts (like IRC channel flooding), or to
send it inappropriate advertisements (and on almost all email lists or
newsgroups, _all_ advertisements are inappropriate).
(16) Filtering blocks unwanted material from getting onto specific news-
groups or email lists. A newsgroup (within Usenet) is a bundle of messages
devoted to a particular topic; the groups occur within a "newsgroup hier-
archy." A moderated newsgroup blocks messages from going directly to the
data-base; instead, they are first read by a moderator, who decides if
they should be publicly posted.
(17) The "@" symbol is used for a number of commands on the MOO, such as
@dig which is the basis for creating a "home" for a user. It has also been
used literary purposes: i.e. "I would @delete you if I could!" Programm-
ing on a MOO uses a MOO language; on some MUDs, there is an LPC programm-
ing language as well.
================================
Bibliography:
Bogucki, Peter, "The Spread of Early Farming in Europe," American
Scientist, Volume 84, No.3, May-June 1996, p. 242
Gurney, O. R., The Hittites, Penguin, New York, 1990
Held, Warren H., Jr., et. al., Beginning Hittite, Slavica, Columbus, 1988
Hodder, Ian, The Domestication of Europe, Structure and Contingency in
Neolithic Societies, Blackwell, Oxford, 1990
Laurel, Brenda, Computers as Theatre, Addison-Wesley, New York, 1993
Malamud, Carl, Exploring the Internet, A Technical Travelogue, Prentice
Hall, New Jersey, 1993
Mitchell, William J., City of Bits, Space, Place, and the Infobahn, MIT,
1995
Postgate, J. N., Early Mesopotamia, Society and Economy at the Dawn of
History, Routledge, New York, 1992
Renfrew, Colin, Archaeology and Language, The Puzzle of Indo-European
Origins, Cambridge, New York, 1987
Renfrew, Colin, and Zubrow, Ezra B. W., The Ancient Mind, Elements of
Cognitive Archaeology, Cambridge, England, 1994
Sondheim, Alan, ed., Being On Line, Lusitania, NY, 1996
Sondheim, Alan, Internet Text, 1994-1996, online at:
http://jefferson.village.virginia.edu/~spoons/internet_txt.html
Sturtevant, Edgar H., A Comparative Grammar of The Hittite Language,
revised edition, Yale, 1951
Wark, McKenzie, Virtual Geography, Living with Global Media Events,
Indiana, 1994
_________________________________________________________________________