First Monday
The Internet Oracle: Virtual Authors and Network Community by David R. Sewell

The Internet Oracle (originally, the Usenet Oracle) was one of the earlier forms of collaborative creation on the Net. The submersion of its authors in an anonymous, collective personality may reflect postmodern notions of the "textualization of the self" in digital writing, or perhaps rather a pre-modern conception of nonprofessional writing as an anonymous dialogue among equals. Participants in the Oracle identify with a "hackish culture" that values wit, self-reflexivity, and anti-authoritarianism, and have managed to create a virtual community with its own conventions and mythos. Originally published in late 1992, this essay has been updated with links to material now available on the Web, and a new Appendix surveys changes in the Internet Oracle since 1992, provides links to new Oracle resources, and considers whether the original conclusions remain valid after four frenetic years of Internet growth.


What Happens to Authors in Cyberspace?
The Internet Oracle: Description and Background
Features Unique to Online Collaboration
Why are Internet Oracle Authors Content to Remain Anonymous?
The Oracle and Textual Authority
The Oracle and the Network Community
Appendix: The Oracle Four and a Half Years Later

What Happens to Authors in Cyberspace?

QuoteMany literary theorists who have addressed the phenomenon of "electronic writing" -- a catch-all category that includes word processing, hypertext, and all communication on wide-area networks -- argue that its immateriality as a medium calls into question the notional status of authors who publish using it. For some, digitized writing is the technology finally responsible for the "death of the author" that Roland Barthes proclaimed two decades ago ("Death of the Author"), a "death" that has been a tenet of poststructural thought ever since. Their basic argument runs something like this: The traditional view of an "author" as a single autonomous agent, the sole intentional creator of a work, is a product of the age of the codex book, when writing was both material and unalterable. But the electronic medium, in Jay Bolter's words, "denies the fixity of the text, and . . . questions the authority of the author" (Writing Space 153). When written words are stored as electronic bits in memory, they are not objects to be owned. When authors are incarnated as electronic texts that can be erased, annotated, downloaded, linked, and redistributed, they are "textualized"; at that point their identities merge into a communal hypertext or discussion thread. Although he wasn't speaking of computers, Barthes had already hinted in that direction by writing that "the metaphor of the Text is that of the network ("Death" 161). Peggy Kamuf finds confirmation of Barthes's formula in the "general incapacity of a conceptual framework to support or contain the author function disseminated by computer-aided modeling and composition, video reproduction, hypertext data banks, nanotechnology, and so forth" (Signature Pieces 16). Perhaps, as Mark Poster suggests, an electronic newsgroup or conference "becomes a single text without an author in the traditional sense of the term" (Mode of Information 122).

Curiously, though, electronic communication has tended to hang on tenaciously to the single, identifiable author: on-line journals have conventional tables of contents and author attributions, nearly all e-mail and news-posting systems identify message senders, and on networks like Usenet the elaborate ".sig" or signature appended to one's postings has become a way of transcending the uniformity of the medium. (A ".sig" is a signature file automatically appended to postings by news software; Usenet posters fill them not only with their name, business or academic affiliation, and e-mail, telephone, and fax information, but also with ".sig quotes" or epigrams, and even fancy ASCII graphics.)

Despite the network's potential to allow anonymous collaboration, it is rare for even experimental network art and participatory projects to be anonymous. For years there have existed on BBS's and conference systems so-called "storyboards" or "never-ending stories," where one person begins a narrative line, and others are free to develop the plot and add characters within the constraints of agreed-upon conventions. (The Usenet group alt.callahans is entirely devoted to communal fiction of this sort, for instance.) Since in most cases contributions to these multiply-authored texts are sent as regular e-mail or news postings, they are identified, sometimes intrusively, as the products of specific authors. Such was the case with "Les Immatèriaux," an experimental e-mail project on which twenty-six French intellectuals collaborated in 1984 as part of an exhibition at the Centre Georges Pompidou. Participant Jacques Derrida observed that while the computer erases the author's voice, each contribution was nevertheless signed and therefore "owned" in the traditional way, making the experiment less radical than the technology allowed. (See Poster 114-115.) Even where an editor intervenes to eliminate tags associating a given author with a portion of the completed text, contributors are typically identified or acknowledged by name. For example, in the collaborative fiction "Thirty Minutes in the Late Afternoon" created and published on the Whole Earth 'Lectronic Link (WELL) in 1990, editor Judy Malloy rearranged posted contributions to create a seamless narrative in parallel columns, but used parenthetical numbers linked to author names to identify the individual contributions.[ 1 ] When claims are made, then, that the electronic medium inherently tends to assimilate the solitary author to a network or collaborative group, they need to be tested critically against the actual practice of writers on the nets.

The Internet Oracle: Description and Background

Yet there are forms of on-line writing which, if taken seriously *as* writing, challenge traditional definitions of authorship because both collaboration and anonymity are enabled or even required by their design. One widely-used example is a real-time conferencing program like the Daedalus Group's "Interchange," which allows pseudonymous login; many writing teachers find that letting students write as make-believe characters can free them to explore a range of voices and attitudes that signed writing might have made threatening. Another example is the Internet Oracle (originally named the Usenet Oracle), in which I have been a participant-observer -- as both anonymous author and named editor. It is one of the most prominent experiments in collaborative creation, an automated mail server that allows two people, a questioner and a respondent, to create a text without knowing one another's identity.

Interestingly, the form of anonymous question-and-respose turns out to be less postmodern in approach than one might expect. It reflects perhaps not so much a postmodern as a premodern approach to authorship, like that of Shakespeare's day, when literature

was a by-product of learning or study, which presupposed leisure. The gentleman might take pride in his by-product, but he considered it as only one of many accomplishments in an active life. He never wrote for money, never put his name on what he wrote, and rarely even condescended to put what he wrote in print. His work was addressed to a small group of equals. (Charvat, Profession 6-7)
The technical basis of the Internet Oracle is a software program developed and installed by Steve Kinzler (with assistance from Ray Moody) on a computer at Indiana University. Although based on earlier programs that ran on local systems, the Internet Oracle was the first to allow any person with e-mail access to the Internet to participate. The concept is simple. A questioner, or "Supplicant," e-mails a question to the Oracle. The Oracle software puts the question at the end of a "question queue"; when its turn comes, it will be mailed to someone else who has submitted a question. That person now becomes an "Incarnation" of the Oracle and must e-mail a response to the question back to the Oracle's address. Finally, the Oracle combines question and answer and mails the completed "Oracularity" to the Supplicant while saving a copy for itself. Because the software encodes all names and addresses, neither questioner nor respondent know one another's identities. [See the Appendix for information on participating in the Oracle.]

In its most basic form, then, the Oracle software essentially automates a party game where a central organizer gathers questions on slips of paper, makes sure that the questioner is not given his or her own submission, and distributes them to be answered. In its early days as a local Indiana University program Oracle was not much more sophisticated than that, its one-line questions and brief answers ranging from witty to flippant:

   The oracle has pondered your question deeply.

   Your question was:

   > Why is a cow?

   And in response, thus spake the oracle:

   } Mu.

   You owe the oracle 2 big kisses.  [000-42]   [ 2 ]

["Mu" is the Zen master's traditional response to an unanswerable question. Questions to the Oracle are always quoted with the ">" character, responses with "}". The original Oracle software automatically appended a randomly-chosen "payment" line to the response; the Usenet version does not, but asking the questioner or "supplicant" for some appropriate payment has become a convention.]

Beginning in October, 1989, when Kinzler publicized the Oracle's existence on a number of Usenet news groups and began posting selections of the best Oracularities to the widely-read rec.humor group, questions and responses became increasingly creative and elaborate, and over the three years of its existence the Oracle has grown far beyond its origins into a genre with its own conventions, formulas, and mythos. The Oracle now has its own Usenet newsgroup,, with current estimated worldwide readership of about 25,000; another 600 subscribers receive digests of Oracularities via an e-mail list [ 3 ].

Two or three digests of ten Oracularities are published most weeks; less than ten percent of all submissions are selected for publication. Beginning with the 100th digest a voting system was introduced so readers could rate Oracularities; every few months the highest-rated ones are collected into a special "best of" collection. After he had read and edited some 20,000 submissions during the Internet Oracle's first year of existence, Kinzler established a "Priesthood" of volunteer editors--currently about two dozen--who filter incoming Oracularities and pass along the best ones.

Not surprisingly, many of the published Oracularities involve computers: parodies of Unix documentation; jibes at Microsoft, DEC, and other companies; elaborate text-adventure games; sessions where the Oracle logs onto a mainframe at "" or onto someone's brain (various parts of the psyche are typically represented as files in a user's home directory); satire aimed at the hated VMS operating system (for instance, a clever comparison between VMS and PMS, which "have pretty much the same features, as anyone familiar with both could tell you" [310-10]). But many other species of Oracularities have evolved: parodies of everything from sociological jargon through pop-cult TV shows to Sam Spade mysteries and 18th-century bawdy drama; humorous and nonsense verse; mock-scientific explanations of obscure phenomena; manic invented histories and science-fiction scenarios. QuoteIn one of the highest-rated Oracularities [135-08], the Oracle runs a simulation program that pits Merlin against Stephen Hawking in order to determine whether magic is real. (Hawking wins when Merlin violates causality by invoking a future self who accidentally kills his earlier self.) The best-received Oracularities are ingenious comic miniatures, often products of considerable effort and imagination. In the following Oracularity, quoted in full, the question sets up an old punch line: "Make me one with everything!" The respondent, however, makes it an occasion for an extended comic monologue:

   The Usenet Oracle has pondered your question deeply.

   Your question was:

   > What did the Tibetian [sic] monk say to the hot dog vendor?

   And in response, thus spake the Oracle:

   } The most famous exchange between a lama and a hot dog vendor

   } occurred one block south of Times Square in July 1988.


   } Hot Dog Vendor: What can I get for ya today?  Footlong with

   } the works?  I said, what can I get for ya today?  Hey, ya

   } wanna hot dog or not?  Listen if yer not going to order willya

   } move on, I gotta business to run.  Stop starin' at me, man.

   } And wipe that silly grin off yer face.  Say something, dammit,

   } yer givin' me the creeps.  Hey, I get it.  Ya don't [speak]

   } English, do ya?  Uh, lessee, yo, uh, tengo los, uh, hot dogs,

   } uh, perros calientes.  Okay, fine! just stand there.  See if I

   } care.  Just don't scare away the customers.  Jeez.  Forget it.

   } Ya wanna Coke?  Coca-cola?  I don't care where yer from, ya

   } gotta understand "Coca-cola".  Coca-cola?  Stop smiling.

   } People'll think yer up to something.  Hey, I got all-beefs,

   } beef-n-porks, turkey dogs, polish sausage, and kielbasa.  You

   } can get ketchup, mustard, mayonnaise, relish, pickles, or

   } onions on them.  I've got plain and whole grain buns.  I don't

   } care what you want, just order something or leave.  I'm

   } serious, man, if you don't go away, I'll call the cops and

   } have them arrest you for loitering.  Jesus Christ, will you

   } stop staring at me!  STOP IT!  At least blink once in a while.

   } You're driving me crazy!  You wanna Coke?  Wait, no, I already

   } tried that.  Listen, man, I'm serious, stop starin' and grin-

   } nin' at me.  I gotta gun under the counter.  I'll use it.  I










   } Then the lama widened his grin just enough to barely show

   } his teeth.  At that moment the hot dog vendor was enlightened.


   } You owe the Oracle a better koan.  And a new deli.

     [293-03; formatting of the original text has been modified]

If this is taken as a representative Oracularity, its most striking features might seem to have nothing to do with "electronic writing." Generically, the response is a dramatic monologue framed as a Zen koan or teaching story. It uses dialect and concrete detail admirably, but no differently than any creative writer would. Even the arbitrariness of the question which the respondent must answer could be mirrored in a traditional writing situation: one can readily imagine a creative writing instructor asking students to compose a scene between a hot-dog vendor and a Tibetan monk as a warm-up class exercise. In fact writing an Oracle response has a good deal in common with impromptu narrative and improvisational drama, two forms that require inventive response to an unforeseen assignment.

Features Unique to Online Collaboration

Other aspects of the Oracle as a writing situation are unique to its medium. Most importantly, questioner and respondent are invisible and unknown to each other. They share neither a physical location nor a common time of writing. Both writers must guess at the likely range of cultural references, terminology, and specific knowledge that their co-authors share. (The Oracle's help file alludes to this problem in its suggestion that writers avoid "slang, jargon, and obscure references," since "[p]eople of all different backgrounds located all over the world use the Oracle.") In the quoted Oracularity, the respondent assumes that readers will understand the humor of the confrontation between New York vernacular culture and Tibetan Buddhism, and that they will catch allusions to the "beatific smile" and the teaching style of Zen (the smiling monk of course plays roshi to the befuddled vendor-novice who is finally enlightened). Mark Poster claims that every author-audience relationship in electronic writing is to some extent a fiction:
[I]ndividuals engage in telecommunications with other individuals . . . without considerations that derive from the presence to the partner of their body, their voice, their sex, many of the markings of their personal history. [They] are in the position of fiction writers who compose themselves as characters in the process of writing. . . . The traces of their embeddedness in culture are restricted to the fact that they are competent to write in a particular language, writing perhaps at the infinite degree. (Poster, Mode of Information 117)
But Poster exaggerates the degree of uncertainty about audience that electronic networks create. Our respondent's assumption that readers would catch the Zen references was not arbitrary. As a discourse community, Usenet has its historical roots in hacker subculture, one of whose best-known features is a predilection for Zen-like paradox; before the Oracularity of the Tibetan monk appeared, there had been many published Oracularities reflecting this interest. (The frequent references to Zen in Douglas Hofstadter's Goedel, Escher, Bach [1979] merely brought to the attention of the wider public an interest that had been part of hacker culture for years. See, for example, the "AI Koans" in Appendix A of Eric Raymond's The New Hacker's Dictionary, pp. 404-405.) Even though Usenet readers are growing far more varied in background as wide-area network use mushrooms, the Net's discourse conventions derive from hacker subculture as surely as the prescriptions of traditional high-school English classes are rooted in neoclassical grammar.

Why does such a talented comic writer as the Master of the Hot Dog Koan choose not to identify himself or herself to gain what we usually consider a major reward of authorship -- recognition? While the Oracle software makes anonymity possible by withholding participants' names and e-mail addresses, the complete anonymity of published Oracularities is actually a matter of convention and authorial choice. In the introductory Oracle help file, participants are explicitly told, "If you do not wish to remain anonymous, you may include a phrase in your answer like "incarnated as <insert your name and/or address here>." Nevertheless, fewer than one percent of authors choose to do so. This was one of the first things about the Oracle that intrigued me: writers like the Master of the Hot Dog Koan were evidently putting real effort into writing that went "unrewarded" by the conventional association of name with publication.

Why are Internet Oracle Authors Content to Remain Anonymous?

In August, 1992, I conducted a survey of Oracle participants to seek answers to this and other questions about the Oracle, receiving via e-mail 125 returned questionnaires. Of the 80 active authors who answered the question "How do you feel about the anonymity of Oracularities?" 59 (79%) felt it was helpful or crucial, while only 5 (7%) said they would prefer to be identified. Narrative responses to the question indicate that anonymity provides two crucial advantages: freedom of self-expression, and the shared aesthetic illusion of an Oracle persona. QuoteLike college professors who publish murder mysteries or romance novels under pseudonyms for fear of being thought unprofessional, Oracle writers sometimes feel safer when unidentified:
I think [anonymity is] essential. I wouldn't have the guts to use the Oracle if I knew my name was going with everything I wrote.

It helps me to give answers which are much more uninhibited. If I knew my identity would be made public I might be a little reluctant to write, since I would not want co-workers to know how much I am involved.

But the second reason for accepting anonymity more resembles that of the medieval author, who, in Hans Robert Jauss's words, wrote "in order to praise and to extend his object, not to express himself or to enhance his personal reputation" (Ede and Lunsford, Singular Texts 78). The "object" in this case is the collection of a corpus of work by a personality, the Oracle, whose characteristics derive from the collective efforts of contributors. (The Oracle help file acknowledges "the thousands of Oracle participants over the years who have created the personality, mythos and history of the Internet Oracle.") And in fact the Oracle has accreted an identifiable personality. Like a Greek god, he is polymorphous: now a crotchety old man, now a super-intelligent computer program, now a deity. A jealous, omniscient and omnipotent being, he is apt to strike with lightning (or "<ZOT>") supplicants who insult him or fail to grovel sufficiently. Nevertheless he is vulnerable to having his plug pulled by his creator Kinzler, his computer's system administrator, or an irate "" Like Zeus, he has a consort: Lisa evolved from the cliche-geek's fantasy-fulfilling "" to the Oracle's companion. It may be that one reason for leaving Oracle submissions unsigned is generic constraint: like Scripture, Oracularities should seem to participants to proceed directly from the voice of God. As E. M. Forster once observed of unsigned newspaper editorials, "anonymous statements have . . . a universal air about them. Absolute truth, the collected wisdom of the universe, seems to be speaking, not the feeble voice of a man" (Anonymity 8). A number of Oracle authors who responded to the questionnaire identified similar reasons for leaving their contributions unsigned:
I'd put less effort into writing for the Oracle if [my identity] were public. I prefer the idea of an all-powerful Oracle rather than the various incarnations scenario. . . . Sometimes it would be nice to say, "I wrote that!" but I prefer to just smile knowingly...

I don't care who wrote it, but it sort of loses something when I see a signature line. Destroys the myth, so to speak.

When I read Oracularities . . . I prefer to think of a faceless deity in a cave somewhere, not I prefer anonymity.

So one of the most powerful conventions governing Oracle responses is the attempt to give voice to the Oracle's persona, a wise but world-weary and sometimes petty deity for whom answering queries is just a 9-to-5 job:

   } Day in, day out, the Oracle hears the cries of despair and

   } ennui that rise from people like you, trapped in an absurd

   } human condition.  "What does it all mean?" you want to know.

   } Time was, a younger and more energetic Oracle tried to answer

   } every existential query individually.  But Usenet has grown

   } apace, and let's face it, "What is reality?" is FAQ number 1.

     [i.e. "Frequently Asked Question"; 
The willingness of Oracle authors to experiment with different voices and personae mirrors something Trent Batson has noticed about network-based writing classrooms: they seem almost "meant for simulation" (Batson, "ENFI" 4) -- that is, for playing with roles, scenarios, and invented characters. Participants seem to agree that the Oracle is in a small way a verbal world constructed by the community. Good Oracularities, one questionnaire respondent wrote, "are necessarily creative and humorous, but I think the very best ones display a sort of *attitude* that the Oracle has. This is hard to define. It's sort of an agreed-upon personality that the collective mind has." Another respondent observed that Usenet in general is "an artifice by which digitheads like ourselves can communicate with each other - it's a really crude precursor to cyberspace, and is a lot of fun. But it's definitely a simulation."

The Oracle and Textual Authority

Because Oracle writing is fluid, improvisatory, and infinitely variable, it tends especially to mock forms of discourse, from computer documentation to scripture, that are formulaic and authoritative at the same time. In Bakhtin's terms Oracularities are thoroughly "heteroglossic," composed of pastiche, parody, fantasy, imitated voices, conventional genres, comic dialogues. If the traditional model of authorship is what Barthes calls the "Author-God" ("Death of the Author" 146), the Oracle undermines it at every opportunity. Parodies of the Bible abound. There is a "Very Strange Version" [391-08], and a marvelously blasphemous version of Biblical history in which (among other things) Jesus is sent to earth to warn man not to ask the Oracle the Woodchuck Question [460-05], and even a dialogue where the Oracle uses a synchronous "talk" program to summon God on behalf of a supplicant:

   } Somebody wants to talk to you, God.


   } >Yes?  Can't you just give them my Internet address?


   } Sure would like to, God.  But you see, in the context of an Oracular

   } message you've degenerated from a halfhearted joke into an

   } irrepressively [sic] boring formulaic answer.


   } >What is this?  What are you saying?


   } God, you're dead.  Not because of that asshole Nietzsche, not because

   } you're old, not because you hang out at the terminal room on Saturday

   } nights.  You're dead because you are invoked for answering questions

   } like "How much would does a wood chuck chuck if a woodchuck could

   } chuck wood?" and "Is Lisa good in bed?"  You're such a stiff. [155-05]

(This dialogue ends when the Oracle deletes his "God" program, casually noting that "God is simply a mathematical construct of mine that I use to amuse myself during spare clock cycles.") Of course, the Oracle's own textual authority is no less vulnerable. If he is an AI program or a computer, he can be undone by system errors, infinite loops, and line noise. (In one Oracularity, the Oracle is on trial for dereliction of duty. Called to the witness stand, "Kinzler" is asked if he knows the defendant. "Yes I do. He's an executable file in my home directory" [238-10].) As a deity, he is beset by Homeric squabbles with other gods and by his own arrogance. As a human male, he is often bested by Lisa in an ongoing battle of the sexes, or held up to ridicule for his thoroughgoing paternalism.

Always, however, the Oracle is a product of writing, and his status as text is often underlined by metafictional play with the Oracle conventions or, more generally, with narrative and language themselves. Asked which types of Oracularities they preferred, respondents to my questionnaire chose as their favorite genre "meta-Oracularities (self-referential plays on Oracle conventions)." Given the preponderance of participants with strong computer backgrounds (78% are, or are preparing to be, "computer professionals"), this is not surprising: of the six distinctive characteristics of "hacker humor" identified by Eric Raymond, the first is "fascination with form-vs.-content jokes, paradoxes, and humor having to do with confusion of metalevels" (New Hacker's Dictionary 203) [ 4 ]. QuoteSome of the best Oracularities play with surrealistic and metafictional frame-breaking in the best tradition of Italo Calvino and John Barth. For instance, in one of readers' all-time favorites, the Oracle is asked who would win a fight between Superman and the Hulk. Finding the question too trivial for his august consideration, the Oracle runs a simulation program on the Vax in search of an answer. Sim-Hulk and Sim-Superman batter each other until the Hulk, about to lose, decides to smash out of the simulation, whereupon he and Superman begin showing up on users' terminals all over Indiana University, and "Kinzler" receives an electronic message from an irate administrator: "Stephen, what is that goddamned Oracle of yours up to now? We have memory faults all over the place, iuvax is threatening to 'smash puny workstations' and this errant process is invading every die green behemoth! You see what I mean? Knock it off! Smash!" [140-05]. Another Oracularity is an extended self-referential tour-de-force along the lines of Barth's "Life Story"; if anything it is more effective in showing narrative giving birth to itself since no identifiable author stands behind the prose:

   The Usenet Oracle has pondered your question deeply.

   Your question was:

   > This is the first sentence of my question, which wants the Oracle to

   > know that all the sentences of my question grovel humbly before the

   > Awesome presence.  . . .


   > Boldly reclaiming the path, this sentence starts out a new and improved

   > paragraph.  This sentence is confident we will finally get to the

   > point, since it can see the next sentence will, indeed, ask the

   > question.  This sentence wants to know if there is anything profound in

   > self-reference. . . .

   And in response, thus spake the Oracle:

   } This is the first sentence of the Oracle's response.  This is the

   } second sentence of that response.  This sentence appears several times.

     [. . .]


   } This is actually the third sentence of the first paragraph but has been

   } placed here in error.  This sentence appears several times.  This

   } sentence attempts to abandon the self-referential style so that your

   } question may be answered, but fails.  This sentence makes the same

   } attempt, but fails just as miserably.  This sentence appears several

   } times.  This sentence, though not able to abandon self-reference,

   } nonetheless succeeds in tackling your question in that it postulates

   } that while the selfreferential style may seem horribly vague and boring,

   } it *does* give ample opportunity for playfulness on the part of the

   } author. [394-10]

The Oracle and the Network Community

The preceding Oracularity is somewhat anomalous in the thorough undermining of authorial presence it borrows from the "high-culture" experimental literature it imitates. Ordinarily, techniques like self-referentiality and frame-breaking in the Oracularities differ subtly from their analogues in literary metafiction. In the latter, they serve to efface the concrete social reality of the author by providing the illusion that the text writes itself. The "author" of literary metafiction is presumed to be the sheer intertextual conjunction of other books, or perhaps an arbitrary language game, like the combinatory that generates the books in Borges' Library of Babel. By claiming origin in pure formal systems, metafiction denies that it is a product of a given society, let alone of an individual author.

However, the Oracle's obsession with logic games, paradox, and infinite regress mark its collective author as a member of a distinctive and identifiable subculture, that of the hacker. Where literary metafiction can be -- perhaps most often is -- apolitical, the Oracle's very existence on the Net is an implicit endorsement of hacker politics: information (both data and text) should flow freely; authority over information systems should be decentralized; the aesthetics of programming (or any other creation; a poem can be a "good hack") is more important than the material uses to which it may be put. Jim Cheetham, one of the Oracle Priests, finds awareness of "hacker culture" important to Oracle participants "because I think it's a good and correct outlook to try to educate 'net users into," especially since the Oracle "has a (slightly) moderated version of the 'net's government by anarchy" (personal communication). The core characteristic of Net governance is that conventions and rules emerge from community practice and consensus rather than being imposed from the top. In many ways the development of Oracle conventions (like those of Usenet writing in general) resembles the evolution of epic in an oral culture: any individual participant is free to alter, supplement, or redirect the narrative, but only those innovations that are accepted by the community survive [ 5 ]. Fascination with recursion is, one might say, sociologically grounded in the Oracle community, as reflected in a wickedly clever Oracular response to the question, "Why don't computer scientists have any sexual stamina?": "Their problem is a fear that any repetitive process is actually the dreaded infinite loop. Providing a proof that the usual termination condition will still occur should suffice" [089-10]. Roger Noe, a computer scientist and Oracular Priest, believes the Oracle's persona reflects "self- satire at its best":

[W]e're really making fun of ourselves, the users of computers, and the designers and implementers of computer hardware and software, which is not necessarily distinct from the group of users. . . . [M]uch of the Oracle mythology is simply a satire of the stereotypical computer nerd. He's a know-it-all who holes up in his sanctum sanctorum, surrounded by every kind of computer hardware and software imaginable, connected to every network that might exist (including olympus-net, god-net, cthulhu-net, you name it) and continuously engaged in multiple simultaneous conversations from people obsequiously seeking his knowledge. (Personal communication) [ 6 ]

In identifying self-satire as a generic motive, Noe helps explain why most Oracularities are *not* couched in terms that only a computer scientist or electrical engineer can understand. A few Oracularities have in fact been written entirely in C programming code (inevitably beginning with the header "#include <stdgrovel.h>"--i.e., an imaginary standard library grovel file), and others that elaborate the syntax of a context-free grammar for Oracle questions and answers, but a heavy concentration of these generally provokes irritation from Oracle participants. As one respondent to my questionnaire, an engineer, put it, "I HATE (!!!!) Oracularities which rely on some intensive knowledge and familiarity with computers, operating systems and languages - computer nerd 'humour' of this sort is pathetic." This even though Oracle authors tend to imagine their audience as composed precisely of computer nerds; as one participant described it, "a room full of sophomore and junior undergraduate computer science geeks with x-rated gifs on their xterms, speaking technobabble to each other, and nary a one of them has had a date or a bath for a month." (As a matter of fact, the average age of Oracle readers turns out to be a comparatively elderly 26.) One motive for not writing technobabble, then, might be to satirize those who can write nothing else. (In his entry on "Hacker Speech Style," Raymond says, "One should use just enough jargon to communicate precisely and identify oneself as a member of the culture; overuse of jargon . . . is considered tacky and the mark of a loser" [New Hacker's Dictionary 20].)

The members of the Oracle Priesthood who answered my questionnaire agreed that a broad cultural knowledge is important to give competent responses to questions. Here is how they rated a number of categories of knowledge on a scale from 1="not important" to 5="very important":

Table 1: Cultural Knowledge Requirements, Oracle PriestHood (scale range: 1="not important" to 5="very important")
Classical (Greek & Roman) mythology3.5
Classic English and American literature3.4
Illuminati, Tolkien, other cult literature3.4
Current world affairs3.3
American popular culture/politics3.2
Geography and history3.1
Hacker culture and lore3.1
Oracle mythos (Lisa, ZOTting, etc.)3.0
Natural and biological sciences2.8
Unix operating system2.8
C programming2.8
Other computer systems (VMS, DOS, etc)2.5

There is a clear continuum here from literature down through current affairs and history to formal and scientific systems. These responses suggest a catholicity, a desire to open the network subculture up to any form of culture it can incorporate. To the extent that it is a hacker phenomenon, the Oracle is the vehicle of a discourse community that is actively assimilating older modes of thought and writing rather than turning inward. So it is no surprise to find, for example, an updated version of the old Davy Crockett backwoods boast:

   > I am a Unix Guru: I debug programs from octal dumps.

   > I eat VMS hackers for lunch.

   > I know the entire Ada manual by rote, never use Ada anyway since I write

   > all my programs in machine language and never use assemblers since I

   > type in the binaries directly using cat. [a Unix program that displays


   [. . .]

   > I write device drivers in my sleep.

   > The DEC salespeople worship me as a minor deity and sacrifice young,

   > buxom secretaries to me at full moon. [141-10]

It wouldn't be too far off, in fact, to call Davy Crockett a nineteenth- century version of the Oracle. The jests, tall tales, and embroidered history published in the popular Davy Crockett almanacs were anonymous vehicles for solidifying the folk culture of the frontier, as the Oracle is for the electronic frontier. The Oracle differs chiefly in its inclusiveness: the electronic frontier is a circle whose circumference is nowhere, and in principal anyone with network access can join the community. Consider the Swedish participant who reported that he appreciates the Oracle because it allowed him to have "a fruitful cooperation with a total stranger, which other people (also strangers) have liked so much as to put [the result] in the oracularities." The evidence is that writing for a virtual community like the Internet Oracle's can be its own reward. "It's wonderful," another participant wrote; "there is an immediate reward in the form of your own question being answered. And if your answer is good, immediate reward by publication."

The existence of the Internet Oracle by itself is hardly enough to herald the "death of the author," especially when so much electronic writing still takes traditional forms. Nor does working with computers diminish individualism simply because networks constitute a kind of communal reality or because programming emphasizes formal systems rather than subjective expression; Eric Raymond notes that in its anonymity "the Oracle is atypical. Most hacker-community projects are undertaken as ego expression, a way to earn the respect and approval of peers" (Personal communication). And the same can be said of most non-hackers who publish on the Net. But the energy and creativity that writers put into their Oracle participation is one among several pieces of evidence that the computer's ability to create self-contained virtual worlds is beginning to affect what we traditionally call "writing" or "literature" as distinct from "mere" game.

Just as the players in 3-D interactive adventure games of the future will "become" knights, hobbits, or New York cabbies, virtual writers in interactive network spaces take on new identities in a universe of discourse where their supposedly "real" selves may never be known, and where even their simulated identities may not be fixed. The Net may yet turn out to be that culture imagined by Michel Foucault "where discourse would circulate without any need for an author . . . [and] would unfold in a pervasive anonymity" ("What Is an Author," 138). In such a culture writers may lose the rewards of traditional authorship, while gaining the satisfaction of helping to create art forms and genres that could not have existed otherwise. As one of the questionnaire respondents wrote, "Oracularities are designed for the medium in which they are read--I can't imagine it working on anything but Usenet." It remains to be seen whether such innovative forms will become, figuratively, the cathedrals of cyberspace that countless unacknowledged builders and designers will collaborate on for the sake of creation itself.


1. "Thirty Minutes" was published in the online journal ART COM, Number 42 (October 1990).

2. All Oracularities quoted in this article were originally published in online digests; the first number refers to the digest volume, while the second is individual Oracularity. The Oracle help file referred to above includes information on retrieving back digests.

3. The readership figure derives from Brian Reid's period estimations of Usenet group readership formerly posted to the newsgroup news.lists.

4. The others are "Elaborate deadpan parodies of large intellectual constructs"; "Jokes that involve screwily precise reasoning from bizarre, ludicrous, or just grossly counter-intuitive premises"; "Fascination with puns and wordplay"; "A fondness for apparently mindless humor with subversive currents of intelligence in it" (Marx Brothers, Monty Python); and "References to the symbol-object antinomies and associated ideas in Zen Buddhism and (less often) Taoism" (203-204). All of these are common in the Oracularities.

5. Asked how Oracle conventions developed, Steve Kinzler replied, "emergently. Someone tries something in an Oracle question or answer, it gets published in the Oracularities, everyone else reads it, catches on and starts using it themselves. The durable ones stick, the weak ones fade away" (personal communication).

6. Another of the highest-rated Oracularities manages to interpret Creation in hackish terms at the same time that it satirizes the stereotyped male hacker:

The Usenet Oracle has pondered your question deeply.
Your question was:
> Why is it that most men suffer a complete loss [o]f personality when
> exposed in any manner to a computer?
And in response, thus spake the Oracle:
} In order to explain this I must detail the story of creation...
} In the beginning there was a Computer. And God said to the computer
} % vi creation.c
} He then wrote the universe, and compiled it and it was good.
} And God ran it in background, and saw that it was good. He
} then noticed that the Universe was eating CPU time and tried
} to kill it, so that he could do his important work, which
} was to determine the Ultimate Question of Life the Universe
} and Everything. The Operating System had a glitch and the
} Universe could not be kill -9'd.
} It came to pass that a lady friend of His wanted to visit
} with Him. He snarled at her for the interruption. Then Man,
} being made in His image, forever duplicated this when being
} interrupted by women while he was working on a computer.
} That is why men react poorly when being interrupted on the
} computer. It is a Divine trait.
} You owe the Oracle the source code for the Universe. [175-10]


Roland Barthes, 1977. "The Death of the Author." In: Image, Music, Text. Translated by Stephen Heath. New York: Hill and Wang.

Trent Batson, 1992. "ENFI and Drama." EnfiLOG 1.1.

Jay David Bolter, 1991. Writing Space: The Computer, Hypertext, and the History of Writing. Hillsdale, N. J.: Lawrence Erlbaum.

William Charvat, 1968. The Profession of Authorship in America, 1800-1870. Edited by Matthew J. Bruccoli. Columbus: Ohio State Univ. Press.

Lisa Ede and Andrea Lunsford, 1990. Singular Texts/Plural Authors: Perspectives on Collaborative Writing. Carbondale: Southern Illinois Univ. Press.

E. M. Forster, 1925. Anonymity: An Enquiry. London: Hogarth.

Michel Foucault, 1977. "What Is an Author?" In: Language, Counter-Memory, Practice: Selected Essays and Interviews. Edited by Donald F. Bouchard. Translated by Donald F. Bouchard and Sherry Simon. Ithaca: Cornell Univ. Press.

Peggy Kamuf, 1988. Signature Pieces: On the Institution of Authorship. Ithaca: Cornell Univ. Press.

Mark Poster, 1990. The Mode of Information: Poststructuralism and Social Context. Chicago: Univ. of Chicago Press.

Eric S. Raymond (editor), 1991. The New Hacker's Dictionary. Cambridge, Mass.: MIT Press.


The Oracle Four and a Half Years Later

What strikes me the most as I reconsider my Oracle piece is how valid most of what I said remains. I don't intend that as self-congratulation: what I mean is that very few generalizations about or analyses of Internet institutions written in 1992 still hold true in 1997, so that the essay's lack of obsolescence is a testimony to the Oracle's stability more than to my insight.

The Oracle has changed, but not much. The biggest difference is its new name: in March 1996, Steve Kinzler announced that the Usenet Oracle would thenceforth be called the Internet Oracle. This change reflected an old reality and a couple of new ones. The Oracle had never been solely a Usenet phenomenon; its basic vehicle of interaction was always e-mail, and many people received the Oracle Digest in their mailboxes rather than by reading the Usenet newsgroup. Moreover, the Oracle had had an official presence on the World Wide Web since early 1995, when Scott Panzer, later aided by myself and other volunteers, created the Internet Oracle Resource Index, which now contains links to all of the Oracle help files, back digests and "best-of" collections, ancillary material about the Oracle, and even Web-based means of submitting questions to the Oracle and voting on the digests.

But the name change also acknowledged the much greater public awareness of "Internet" than "Usenet" in 1996. Consequently, some long-time participants felt it was a cynical quasi-marketing ploy, and rather bitterly denounced the shift on the discussion group as a betrayal of tradition. Oracular tradition had already come under siege during the previous year or two as a consequence of the Invasion of the Masses, the huge influx of non-academic, non-research-community Internet users who began to arrive in 1993 as large commercial ISPs added access to Usenet groups, and who by 1996 had permanently changed Usenet, mostly for the worse, with their ignorance or defiance of traditional norms, conventions, and "Netiquette."

For a while, in fact, it seemed that the Oracle might not survive as the core of a discourse community. The average quality of responses fell off drastically: where Kinzler had originally expected that about 10% of Oracularities would be good enough to share in the Digest, the proportion has declined to something more on the order of 1-2%. Supplicants could no longer expect as a matter of course to have questions answered by Incarnations who shared the same general body of knowledge, values, and literacy. Frequent complaints were posted to about responses that were ignorant, abusive, or both, a state of affairs satirized in mid-1995 in Oracularity 760-06, for instance.

To swipe a hoary Net meme, the Imminent Death of the Oracle didn't happen. New participants gradually assimilated to its conventions, or wandered off to experiment with other interactive phenomena on the Net. (A glance at Yahoo's category "Add-To Stories" suggests how common these have become on the Web, something that undoubtedly dilutes the impact of newcomers on any single institution like the Oracle.) In addition, the Usenet group, passed over unmentioned in my original essay, has become a much more important virtual community and is arguably now the most powerful force in maintaining the traditions and norms of the Oracle. And in extending them, for the Oracle mythos has not stagnated. In early 1995, prolific new contributor Richard Wilson introduced Zadoc the Priest, an ultra-obsequious but occasionally subversive companion of the Oracle who quickly assumed a prominent role as schlemiel among the Oracular dramatis personae. (Most recently, Zadoc contributed a rare addition to the communally accepted "plot" of the Oracle's biography by running off with Lisa.)

From the perspective of 1997, the one part of my original analysis that I would modify is the claim that as an anonymous collaboration the Oracle confirms poststructuralist ideas about the dissolution of the self into digital textuality. The commercialization of the Internet has shown the weakness of the earlier theorists' equation of cyberspace / hypertext / "the electronic word" with postmodern discourse. Although the first corporations to venture onto the WWW sometimes emulated the then-trendy cyberpunk sensibility of the early '90s, the Time-Warners and CNNs who have massively colonized Webspace are rapidly turning it into an extension of a monolithic information space that tends to look the same whether incarnated as print, television, or HTML. Clearly, the mere fact that words and graphics are digitized does not entail that they follow any particular aesthetic or politics. The Internet today is much more like a palimpsest of many historical and generic conventions than a uniquely postmodern space.

So I think I was closest to the mark in equating Oracle authorship with Charvat's picture of the gentleman-writer who addressed his work to "a small group of equals". Although anonymity in the published Oracularities remains as much a convention as ever, it has become common for the more active and skilled participants to identify themselves as authors of particular contributions on, and to "publish" collections of their own work via links from the Oracle Resource Index. What motivates and shapes the Oracle, then, seems to me now less a function of some generalized notion of digitized writing than of a particular discourse subculture, much like (say) that of storytelling festivals or Star Trek fan fiction. Certainly the Oracle remains profoundly shaped by its origins in particular computer technology and culture--the "hackish" identification of participants remains high even as the proportion of them who are computer professionals has surely declined (witness the steady flow of Oracularities ridiculing Microsoft). But I see more continuities now between the Oracle and pre-digital modes of writing than I perhaps did in 1992--and maybe after another decade or so of Oracularities I'll be able to decide which perception is the closest to the truth.


Special thanks to Steve Kinzler and all of the Oracular Priests and participants who corresponded with me as I was doing research for this article. This essay originally appeared in EJournal, an online peer-reviewed publication, where it carried this notice:
This essay in Volume 2 Number 5 of EJournal (December (2), 1992) is (c) copyright EJournal. Permission is hereby granted to give it away. EJournal hereby assigns any and all financial interest to David Sewell. This note must accompany all copies of this text.End of article

The Author

David Sewell has read and contributed to Usenet since 1981, when he had his first Unix account as a graduate student at the University of California, San Diego. He was formerly a member of the English Department at the University of Rochester, where he helped establish a Humanities Computing Center and explored the use of collaborative hypertext in writing instruction. Currently he is a scientific editor at the University of Arizona, and is also a contributing columnist to the "Going Meta" section of Go2net ( Contact: E-mail: Telephone: +1 520 881 0857 Post: c/o Geosciences Department, University of Arizona, Gould-Simpson 208, Tucson, AZ 85721 USA

Copyright © 1997, First Monday

The Internet Oracle: Virtual Authors and Network Community by David R. Sewell
First Monday, Volume 2, Number 6 - 2 June 1997