Formal and Informal Rules

The future manager/owner of the IBMPC computer conferencing facility recognized, soon after the opening of IBMPC in 1981, that the future of the facility depended on the support of IBM management and the existence and enforcement of formal rules that would:

Hence creation of an initial set of formal rules for IBMPC became the first priority of Dave Chess when Dr. Waldbaum assumed ownership of IBMPC in early December, 1981. The first document containing such rules, IBMPC MEMO, was written by Chess under Waldbaum's guidance and was distributed just a few days after they assumed control of the facility. IBMPC MEMO was principally intended as a user's guide to the facility. It stated what the IBMPC computer conferencing facility was for (sharing information about the use of IBM Personal Computers), how one could access information on the facility, and how one could contribute information to the facility. The rules stated in the document were few:

The following rules are suggested for material which is appropriate to the IBMPC disk. They will be enforced by the owner of the disk (hopefully not at all).
  1. Material must be related to the IBM PC, either:
    1. things that run on the PC
    2. things that run on 370 to produce stuff for the PC
    3. information or requests for information about the PC
  2. Material must be public domain or IBM internal use only (not confidential).
  3. Material should not have patent or copyright problems.
  4. Material must not cause system problems at Yorktown.
  5. Material must be well documented and reasonably bug free.
  6. The size of things should be constrained. (send a MEMO that points to large objects)
  7. Look at other available commands and do not duplicate function with minor changes. Work with the owner of the command you would like to extend.

Only one of these rules, the last, deals in any way with the nature of interaction on IBMPC, directing users to work together. The rest of the rules are concerned primarily with protecting IBM's property, ensuring there are no legal exposures, insuring usability, constraining the amount of mainframe disk space the facility consumed, and insuring that participants understood what the facility was to be used for. None of these rules originate in the experience of computer conferencing. They are based, rather, on a variety of existing rules that were already documented in various IBM documents, including the managers manual, VNET rules, and security guidelines.

These rules represent, in many ways, a starting point in the process of defining IBMPC as a computer conferencing facility. Indeed, one notes without surprise that only the first three of these rules survive today in any form (and they haven't survived intact). The ability to directly interact through the facility was introduced at the same time that IBMPC MEMO was. Hence the relative paucity of interaction rules reflects the simple fact that no one knew what rules for interaction ought to exist. This rather brief emumeration of IBMPC's first formal rules is hardly the sum of the rules governing interaction on IBMPC, however.

Relevant company rules

As IBM employees, participants were accustomed to operating within the "IBM culture", an unstated set of norms which still reflects, to some extent, Thomas J. Watson's injunctions against drinking, casual dress, and rude behavior. They were expected to behave within the constraints of IBM's three basic beliefs and Business Conduct Guidelines, which set high standards for employee behavior in many areas, starting with "respect for the individual" (IBM, 1988, p. 10). Finally, they had to contend with the preferences of the broad consensus of other IBM employees.

The behavior of IBMPC participants (with its presumed basis in the pre-existing rules and cultural norms of IBM) seems to set IBMPC apart from other computer conferencing facilities. This seems especially true in the wake of Keisler, Seigel, and McGuire's (1984) description of flaming and other computer conferencing behavior. These descriptions painted computer conferencing as a medium where unrestrained language was the norm. Early appenders of SENSITIV FORUM expressed IBMPC's contrasting behavioral norms clearly:

The origins of today's formal IBMPC rules

Hence when the administrators of IBMPC substantially revised the formal rules that govern interaction on the facility in mid-1984, it was not the behavior of IBMPC participants that drove the effort. It was, rather, the growth of conferencing within IBM. IBMPC was still the only computer conferencing facility in the company, but it was becoming clear that there would soon be other conferencing facilities. Indeed, the forums that would eventually seed IBMVM and other facilities were already operating on IBMPC and discussions about the possible opening of IBMVM were starting.

What was needed, in the judgement of IBMPC's owner, was a set of formal corporate computer conferencing guidelines that would govern the conduct of computer conferencing both on IBMPC and on any future computer conferencing facilities. IBMPC's existing formal rules, still located in IBMPC MEMO, were probably adequate to the continuing administration of IBMPC, especially given the range of informal understandings that had evolved around them. Formal corporate guidelines, it was felt, would require a more detailed statement of the rules that made those informal understandings, which had evolved over several years of running IBMPC, clear.

Hence the 1984 revision of the IBMPC rules was intended, more than anything, to clarify the existing IBMPC rules so that the owners of new conferencing facilities could understand them fully while providing a base from which corporate wide computer conferencing rules could be drafted. The concern here was that if another computer conferencing facility was not appropriately managed, the problems of that facility might also jeopardize IBMPC. Hence, adoption of these revised formal IBMPC rules by IBM's second computer conference, IBMVM, was a precondition to the opening of that facility.

Influences on the formal rules

These rules, both as they were initially drafted and as they have subsequently evolved, are shaped by two major influences. The first was IBM's existing formal corporate policies and guidelines, which acted directly in providing impetus for some rules and indirectly in shaping the formal and informal rules the facility was already observing. A second influence was found in the informal rules that already governed the behavior of IBMPC participants.

The nature of these influences is strikingly different. IBM's formal policies and guidelines provide a formal standard against which IBM can measure the IBMPC rules. If IBMPC's rules are the metaphorical equivalent of state laws, the IBM policies and guidelines are federal laws which supercede the IBMPC rules. Strictly speaking, IBMPC's rules (both its stated formal rules and its informal rules) must conform to relevant IBM policies and guidelines.

IBMPC's informal rules, by contrast, represent the IBMPC participant's general consensus on the kinds of conferencing behavior that maximise the usefulness of the medium. There are many informal rules guiding behavior on IBMPC. A number of these rules will be discussed later in this chapter. Most informal rules stay informal. They are not really stated as rules anywhere but in the knowledge and behavior of IBMPC participants. Some such rules are of sufficient importance to the conduct of computer conferencing on IBMPC that they have been turned into formal rules.

The difference between the sources is one of control and influence. IBM's existing policies and guidelines control IBMPC's formal rules by setting a baseline that the IBMPC rules must conform to. IBMPC's informal rules, by contrast, suggest constraints on conferencing behavior that, although not specifically addressed by existing corporate rules, policies, and guidelines, are important to the success of conferencing. Both sources of rules are important. The IBMPC manager owner and his team have attempted to insure that rules of IBMPC serve the needs of IBM, continually updating them where needed to address new corporate requirements and the ever-evolving changes in the way IBMPC is used. In the end, then, IBMPC participants can influence the rules, but the owner manager of the facility has the final responsibility for deciding what rules (both formal and informal) will be accepted and enforced.

The current IBMPC rules

As written today, the formal rules of IBMPC stipulate that IBMPC's purpose is to facilitate the management-approved exchange of information directly or indirectly related to IBM's various personal computer products. According to these rules, contributions to IBMPC:

Measured by their original goals, IBMPC's current formal rules have been judged to be successful. The rules have been adopted, generally with very few changes, by nearly every computer conferencing facility within IBM. When, moreover, IBM issued formal corporate guidelines for computer conferencing within the company in 1988 (more on this later in the chapter), those guidelines strongly reflected the formal rules of IBMPC. Indeed, the guidelines were based so heavily on existing practice that their release had almost no effect on the conduct of computer conferencing within IBM. The only major change from what existed before was the formal specification of the agreement contributors of software needed to agree to when sharing their personal efforts with the rest of the company.

Specialized formal rules

Additional formal rules are associated with some of the more specialized forums on IBMPC. One such forum, NEWSCLIP, is a real-time newsletter in which volunteer "reporters" provide abstracts of current personal computer and IBM business related technical news. The special rules NEWSCLIP has evolved include:

The evolution of special rules for forums like NEWSCLIP reflects the desire of many IBMPC participants to adapt IBMPC to special purposes. It is, for instance, the long standing experience of IBMPC participants and administrators that some news items can be expected to provoke intense debate about such things as company strategy, product quality, and software development methods. These and other recurring issues frequently have no "right" answers, but frequently evoke strong opinions. The forum-specific rules associated with NEWSCLIP exist primarily to protect the forum from such debate and ensure its productivity. Most of NEWSCLIP's special rules are intended to discourage all discussion on the forum; to remove its interactive quality; to make it effective as a "mass medium".

Even with such rules, NEWSCLIP has proven to be a particularly difficult experiment. The forum's owner argues, citing personal mail from forum readers, that the experiment has been a successful one, but largely because of strong editorial guidance which orients NEWSCLIP to readers rather than contributors. Even with well publicized rules, a general consensus of IBMPC participants that the rules are reasonable, a resultant peer pressure to conform, front end software that tries to prevent problems before they occur, and a strong editor who deletes off-topic appends quickly, IBMPC participants persist in occasionally attempting to start discussions on NEWSCLIP and in complaining about editorial action when it is taken.

Other elements of the NEWSCLIP experiment have been considerably less successful, and the owner of NEWSCLIP continues to work with the IBMPC owner and administrators management to find ways to improve its workings. The experience of NEWSCLIP raises a question of how the structure of a medium affects the rules that can be imposed on it. It appears to be the case, at least on the IBMPC computer conferencing facility, that there are some forum rules which simply cannot be successful without rigorous enforcement by the forum owner. This may be particularly true for forums like NEWSCLIP whose operation and rules differ somewhat from those of other forums. NEWSCLIP looks like a regular forum to anyone encountering it for the first time, and some level of inadvertent violation of its specialized rules will probably be inevitable so long as this remains the case.

The Lightning Rod theory

The problems associated with NEWSCLIP are not new. Indeed, they date back to, at least, SOAPBOX FORUM and its driving hypothesis, the idea of the lightning rod forum. The underlying theory of the "lightning rod forum" is a simple one. Forums, like conversations, frequently are tangential. An odd side statement in an append that is generally related to the subject matter of the forum may incite an extended and elaborate side discussion. These side discussions are often fascinating. They can also be quite deadly to other discussion. When carried to an extreme, such side discussion has the potential to choke off the principle discussion of a forum and to chase away its regular readers.

The "lightning rod" forum exists for the specific purpose of attracting such discussion away from other forums. The subject matter for such forums is, in consequence, stated quite broadly. A prototype statement of the subject matter of such forums might read as follows: "This forum exists for the discussion of a specified range of IBM Business Related PC topics which don't really belong on other forums." The range of topics which are appropriate to such forums are many, and include narrow topics which don't merit their own forums, transient topics which are interesting only for a brief time, and new ideas that may develop to the point where they require a forum of their own, but aren't sufficiently developed yet.

The idea of the lightning rod forum starts not long after the opening of SENSITIV FORUM in October, 1984. SENSITIV FORUM was not designed as a lightning rod forum. It was really an attempt to provide an avenue for meta-communication; a place where IBMPC participants could use computer conferencing to talk about the experience of computer conferencing. Its explicit purpose was and is to provide a forum for the discussion of miscommunication and misunderstanding in computer conferencing, which appeared to be a substantial problem at the time. It was intended to give IBMPC participants, contributors and readers alike, a chance to talk about how such misunderstanding could be reduced. SENSITIV FORUM can be regarded as a success. Many useful meta-discussions of computer conferencing behavior have been held on SENSITIV FORUM, which will be an object of scrutiny later in this chapter.

The opening of SENSITIV FORUM had an interesting biproduct. As SENSITIV FORUM took over part of the responsibility for meta-discussions of conferencing behavior, similar discussions on other forums (where such discussion was off topic) declined. Indeed, participants who attempted to start such discussions were frequently directed (by other participants) to move their opinion to SENSITIV FORUM. The effect of the forum was, at least in part, that of a lightning rod, pulling off topic discussion away from other forums. It was in recognizing this effect that the "lightning rod" theory was born.

The first explicit attempt to apply the lightning rod theory in a more general way came with the opening of SOAPBOX FORUM in April, 1985. The mission of SOAPBOX was really quite simple. It was to be the home of any and all PC-related discussions that didn't belong on other forums; a designated repository for all sorts of tangential discussion. It was to draw such tangential discussions away from other forums, where they would be considered noise.

By some measures, including this basic mission statement, SOAPBOX must be considered a success. It did, as was intended, become the home of all sorts of discussions that didn't belong on other forums. It appeared, moreover, especially in the beginning, that many forums benefited from a reduced tendency to tangential discussion. Another benefit was the larger audience it attracted to computer conferencing.

But the benefits of SOAPBOX were balanced by other factors. With its rather loose topic guidelines, SOAPBOX grew at an (for the time) unbelievable rate, frequently attracting 100 or more appends in a day. Volume is not, of course, inherently bad. There are many forums on IBMPC today that attract more than 100 appends in a day from time to time without creating any problems at all. Appends (and whole lines of discussion) to SOAPBOX had a much greater than average tendency, however, to create problems. These problems seemed to emerge from several sources:

Each of these sources created problems for the administration of IBMPC. These factors led to three general problems. First, they created a substantial workload for the forum owner and IBMPC administrators, who frequently had to delete appends from the forum. Because there were many attempts to start discussions that were neither PC nor business related, there were many such discussions to clean up. Because, moreover, the more controversial of these discussions tended to grow quickly, this clean up frequently involved the deletion of many appends (a lot of work) and a certain amount of meta-discussion (discussed at length in the next chapter).

Second, the existence of such discussions, even though they were deleted quickly, led to a perception by some IBMPC participants that literally anything could be discussed on IBMPC. There was, as a result, a general spillover of the problems of SOAPBOX FORUM to other forums. This spillover presented the IBMPC administrators with a major problem. SOAPBOX FORUM was intended to solve problems by attracting problematic discussion away from other forums. Its effect, however, was sometimes the reverse. It was actually creating new problems on other forums.

In plain terms, SOAPBOX became a lot of work for the administrators of IBMPC. As the facility grew, SOAPBOX grew. Eventually, especially as formal rules and reviewing came into play, the forum's problems became too difficult to manage. Several attempts to manage the forum failed. SOAPRULE FORUM discussed and eventually worked out a set of special rules for the forum. The introduction of these rules, the first specialized forum rules on IBMPC, did not help, and a stronger solution was sought in the renaming of the forum (an action which will be discussed in greater detail).

A shell of SOAPBOX FORUM survives today under the name TEMPMISC FORUM. The forum can, according to its first entry (the "header" in the vocabulary of IBMPC) "be used to discuss any topic which relates to both the IBMPC Personal Computer and IBM business." Discussion is severely constrained, however, by its central rule: "The conversation would be on topics of temporary, short term value, since the forum will be pruned on a regular basis. If the topic is of such value that it should be preserved, it should be moved to a forum, or a forum may be started if one does not already exist." Unless volume is particularly small, appends are usually deleted from TEMPMISC on a weekly basis.

The effect of the name change and regular pruning has been to transform TEMPMISC into a fairly slow growing forum that occasionally gives birth to new forums. It is no longer a "lightning rod" forum drawing tangential discussion away from other forums so much as a nursery where topics that don't quite fit anywhere else can get started.

One should not conclude from the experience of SOAPBOX that the lightning rod theory is entirely without merit. SOAPBOX and its successors, including TEMPMISC and NEWSCHAT, have been fairly successful in drawing tangential discussion away from other forums, much as they were intended to do. It is clear, however, that lightning rods must be installed carefully, lest they draw more lightning than they were intended to. Specialized forum rules and considered naming appear to be key elements of this installation.

Informal rules

There are many aspects of the interaction on IBMPC which are not, and should not be, the subject of IBMPC's formal rules. Indeed, the broad majority of rules that constrain behavior on IBMPC remain informal rules that are enforced by the general consensus of the IBMPC community. These informal rules are not stated all together in one place. They do not, in any case, supercede the formal rules of IBM or IBMPC. They exist, nonetheless, both in the example set by IBMPC participants and in the peer pressure that is brought to bear on offenders when violated. Hence:

The above reflects a sampling of some of the better-known informal rules that constrain behavior on IBMPC. They are, for the most part, rules that make IBMPC easier to use. Indeed, the primary driving force in the evolution of many of IBMPC's informal rules is probably "information overload". Several of the rules outlined above reflect the desire of IBMPC participants to be as productive as possible in their reading of forum content. Most informal rules reflect the consensus of IBMPC users that particular conferencing behaviors are unproductive for the IBMPC conferencing community.

The evolution of formal and informal rules on IBMPC is, as it is for all human communication processes, ongoing. IBM continues to refine its perception of what kinds of things should and should not be discussed on an IBM Internal Use Only computer conferecing disk. IBMPC participants continue to find forms of conferencing behavior that, in their assessment, need to be discouraged or encouraged in the interests of making IBMPC easier to read and more functional.

Murder by meta-discussion

The business of discussing the informal rules of IBMPC is reserved to a class of forums which called meta-forums, forums dedicated to the discussion of discussion. There are a number of meta-forums on IBMPC, including IBMPC FORUM, REVIEWNG FORUM, SENSITIV FORUM, PCTOOLS FORUM, and others. The intent of these forums is to provide a home for meta-discussions that would be considered clutter if they occurred on other forums.

The need for meta-forums arises from a single persistent characteristic of such meta-discussions: the difficulty people have in reaching agreement in them. Meta-discussions on IBMPC come in many flavors. One of the more persistent variants starts with the question "Why is this append (or forum) on IBMPC?" The question, when asked directly on IBMPC, most generally expresses a productivity concern along the lines of "The existence of this type of append (or forum) makes it harder for people to use IBMPC productively."

The immediate problem in this type of meta-discussion is the difficulty in reaching agreement on what exactly constitutes a business-related append or forum. Consider, for instance, a discussion of the style of interaction in a given game as might be observed in GAME FORUM. A critic of this discussion might note (and not without some justification) that game playing is hardly a business use of a computer. This critic might easily argue that the discussion cannot possibly help IBM employees do their job and that the discussion represents a personal use of IBMPC that should not be allowed under the IBMPC rules.

Defenders of such discussion might counter (and not without some justification) that understanding the user interface and algorithmic features of today's game software is an important step toward understanding the user interface and algorithmic requirements of tomorrow's business software. Computer games are frequently industry leading in exploiting the features of competitive hardware and exploring user interfaces, interaction styles, presentation graphics, artificial intelligence techniques and other elements of business software. Hence it can be argued that discussion of computer games is clearly business related.

The problem is that, however well justified each may be, the perspectives are extremely difficult to justify to each other. Winning and losing in this kind of debate does not turn on being right or wrong. It turns on who gets to make the definition of what is business related and how they cast the definition. Hence it should hardly be surprising that this kind of discussion can rapidly escalate into an intense debate encompassing thousands of lines of text.

This kind of meta-discussion was observed a number of times on the technical forums of IBMPC in its early years, generally with highly predictable results:

  1. no action is taken on the issue raised in the meta-discussion. The opinion of the IBMPC owner and reviewer concerning the appropriateness of a given append or forum has generally been decided early in the discussion. Indeed, particularly in cases where a forum's existence is challenged, it is generally the case that the issues raised have been considered in detail long before the related meta-discussion ever gets started. The IBMPC administrators, acting out of a general belief that participants who understand a given decision will do a better job of observing it, may participate in such discussion, especially in its early stages, with the intent of providing such understanding.
  2. an attempt is made to end the discussion. If the discussion swings out of control and threatens the regular content of the forum, the IBMPC reviewer will often ask appenders to end the discussion. This request is frequently successful, but when it is not ...
  3. the patient dies. The volume of meta-discussion on the forum overwhelms the usually much smaller volume of technical discussion had been occurring on the forum. Participants become frustrated with their inability to hold useful technical discussions on the forum, and stop reading it.
  4. the question is not resolved. In the best case, discussion ends with an agreement to disagree. In the worst case, the forum is locked by the forum owner or IBMPC reviewer, effectively stopping all discussion in the forum.

The overall effect of the meta-discussion, as described here, is problematic. If we must count winners and losers, both sides win and both sides lose. The critic wins and the defender loses when the forum dies. The defender wins and the critic loses when the perspective that started the forum is not repudiated. In the end, however, it is the users of the forum that lose, as large number of former participants stop participating.

Recognition of this syndrome, which might be referred to as "murder by meta-discussion", led IBMPC's administrators to formally ban meta-discussion from IBMPC's technical forums in 1984 with a rule that has evolved to read:

If you think that a given item should not be allowed on IBMPC, please SEND MAIL TO THE IBMPC FOLKS, OR TO THE AUTHOR OF THE ITEM, rather than putting such comments on IBMPC itself. Such meta-discussions tend to raise arguments, waste disk space, etc., and will be prime candidates for removal themselves. This also applies to applications of the rules that seem incorrect; if, for instance, you disagree with some editing action by a file owner or reviewer, communicate that disagreement by SENDING MAIL to the editor, *not* by starting a brawl on IBMPC. All appendable files are subject to being purged of such meta-discussions at the discretion of the owner or reviewer.

Most meta-discussion on IBMPC never comes close to the extreme described here, which has much in common with polar debate as described in the discussion of genre. Where the extreme occurs, the forum owner or IBMPC reviewer can be expected to close down the discussion. Even when the meta-discussion is brief and well behaved, however, it is unusual for a meta-discussion to have any real relavance to the technical discussion that is the focus of most technical forums. Hence, based only on such formal IBMPC rules as the requirement that contributions to IBMPC be PC related, business related, and help employees do their job, there would be good reasons for banning meta-discussion from IBMPC even if it did not sometimes lead to "murder by meta-discussion."

The meta-forum

Such a ban is not entirely functional, however, and it has proven much easier to state a ban on meta-discussion than it is to enforce it. Disagreements still occur. Meta-discussion still ensues, sometimes on forums and sometimes in private notes with long copy lists. Eventually an attempt was made to resolve the problem with a new forum, SENSITIV FORUM, which opens with the following statement:

There seems to be a remarkable tendency for people to take offense, in one way or another, to things that were never meant to offend. Jokes seem to be particularly prime candidates for misreading, as are critiques of hardware and software and, most particularly, critiques of critiques.

The problem, according to the forum's third append, is the absence of "the vast amounts of communication that takes place over the human non-verbal channels: the eye wink, the body stance, the flicker of a smile. They all modify the spoken word and avoid many of the difficulties which arise in computer conferences." A solution, according to this same append, might be found in the use of "meta-emotion words" like "sigh", "gasp", and "hehe" Subsequent appends extend the idea and relate it to other media:

An informal consensus of IBMPC users formed fairly quickly around the use of single line character icons, in part because, as a Kingston appender noted, "it takes only a single line of text, and thus is easy to type in the middle of a note. The :-) icon (and variants including :-> and :^)) are frequently used on IBMPC as indicators of a (usually subtle) joke. The :-( icon indicates unhappiness with something. A wide range of similar icons (most of them with less globally understood meanings) has also evolved, including ";-)" (a wink), 8-) (wide eyed glee) and others. Adoption of "meta-emotion icons" as an informal convention is now widespread on IBMPC and many other conferencing facilities inside and outside IBM.

One cannot, in fairness, attribute the convention to IBMPC. The appenders from Raleigh and Kingston who almost simultaneously suggest the convention both claim to have seen the :-) icon used elsewhere. Still, thanks at least in part to the discussion on SENSITIV FORUM, the convention took. Iconic substitutes for nonverbal communication established SENSITIV FORUM as a place where the informal rules and practices of IBMPC could be negotiated. A variety of iconic substitutes for nonverbal meaning established on the forum continue to be used routinely as an indicator of mutually agreed-to meanings. Dictionaries of such icons (some in the form of programs) now show their agreed on meanings to anyone who takes the trouble to look, and there is a general consensus that they work. Insofar as the use of meta-emotion icons acts to clarify intent, it acts to reduce the probability of misunderstanding.

The ongoing need for meta-forums

If character icons were sufficient to solve all problems of misunderstanding on IBMPC, SENSITIV FORUM might have been retired as a repository for "meta-emotion" icons after its first two weeks. If formal rules were sufficient to solve all the problems of disagreement on IBMPC, there might have been no need for new rules after the ban on meta-discussion in technical forums was implemented. If, moreover, misunderstanding and disagreement were the sum of the problems that sometimes need to be resolved when people communicate via computer conferencing, there would be no need on IBMPC for other forums dedicated to meta-communication concerning the conduct of IBMPC.

Such is not the case. There seem to be many issues related to the way people communicate on IBMPC that sometimes need to be resolved. This is, perhaps, as it should be. The process of human communication is, in part, a continuous renegotiation of the rules of conduct, and meta-communication is the principle means of such negotiation. The problems of murder by meta-discussion made a ban of such meta-discussion a requirement for the reasonable functioning of technical forums on IBMPC. But the existence of such meta-discussion to begin with was symptomatic of an ongoing need for meta-discussion. The meta-forum is the IBMPC computer conferencing facility's solution to this paradoxical requirement to both restrict and encourage meta-discussion.

A range of meta-forums

There are a fairly large number of meta-forums on IBMPC (at least 15 at last count). Each attempts to satisfy a subset of the meta-communication needs of the IBMPC community. The following forums can be counted among the most important:


REVIEWNG FORUM is the designated repository for discussions of the the reviewing actions of the IBMPC administrators. It was started by John Alvord, then an IBMPC administrator, with the intent of explaining the reviewing actions taken by the IBMPC administrators. Its focus is the actions of the IBMPC conference administrators, and its most persistent subject matter is the question of what does and does not belong on the facility. This forum is, in some sense, the designated home for the kind of problematic meta-discussion that has been described above, and its subject matter frequently includes discussions on such issues as:

REVIEWNG FORUM is, with SENSITIV FORUM, one of the two primary vehicles for discussion of the interpretation of IBMPC's rules and the actions taken by its administrators. The meta-discussions on REVIEWNG FORUM are almost always specific to a particular event. A given action should or should not have been taken. A given append was deleted and shouldn't have been. A given append was not deleted and should have been. A given forum ought to be allowed on IBMPC. A given forum ought not to be allowed on IBMPC. Its subject matter, in general, is the evolving borderlines of IBMPC's formal rules. It is concerned, by and large, with matters of how the IBMPC rules are, and ought to be, interpretated.

As might be expected in such discussion, the content of REVIEWNG frequently turn on definitions and boundary conditions that will be discussed in greater detail in a coming chapter. When is something X rather than not X? What does it mean to be X or not X? Is a given forum or append X or not X? These are generally controversial issues that people can never quite agree on, and discussions on REVIEWNG FORUM sometimes grow quickly. As might be expected in discussions that turn on definition and interpretation, the problems discussed are rarely resolved to everyone's satisfaction.

It appears, however, that there is a satisfaction in being able to express one's outrage, and participants in REVIEWNG FORUM often express satisfaction with such expression even when the issue in question is not resolved to their satisfaction. It is often the case, moreover, that discussions on REVIEWNG FORUM foster better understandings on the part of both contributor and administrator. IBMPC's administrators often learn things that they didn't know when a particular reviewing decision was made, and sometimes reverse decisions on the basis of that information. Contributors most often come to understand why a particular decision was made and the difficulties entailed in making it. This understanding is extremely important, as it reduces the likelihood that a particular problem will recur.


SENSITIV FORUM is a repository for discussions of the problem of misreading on IBMPC. It is intended to provide a home for discussions of how misunderstandings arise on IBMPC, and the things we can do to reduce such misunderstanding. Most discussions start with a general declaration about how people should or should not behave in making appends to IBMPC. The problem, more often than not, is one of reasonableness:

Informal rules and meta-forums

SENSITIV FORUM is similar to REVIEWNG FORUM in many ways. Both attract complaints about the way IBMPC operates and discussions about how it might operate better. Both act as focal points for discussions of the rules that govern or should govern behavior on IBMPC. If there is a difference between the forums, it is the kind of rules that are discussed.

REVIEWNG FORUM's major concern is the interpretation and application of IBMPC's formal rules. Its discussions focus on boundary conditions, grey areas, and matters of definition. SENSITIV FORUM, by contrast, is concerned mostly with IBMPC's informal rules. Its appends are mostly concerned with the reasonableness of conference behavior which is not covered by IBMPC's formal rules. Discussions on SENSITIV FORUM will, like those on REVIEWNG FORUM, often turn on questions of definition, but there is less expectation that an IBMPC administrator will eventually designate a given definitional position as "correct".

If there is a dominant pattern to SENSITIV FORUM discussions, it is that of an appeal to reason. An appender will present some persistent conference behavior that they find abhorrent or frustrating. Other appenders will discuss the behavior. Those that disagree will most frequently point out conditions that make the frustrating behavior understandable and even acceptable. Those that agree (on either side) work to refine the argument in the face of criticism. While there is rarely a clear cut closure on the discussion (an append stating that "I guess we all agree that ..."), discussion most often ends in one of three ways:

  1. a general consensus that a given behavior is probably a good or bad thing. Where it exists, this consensus usually exists as a series of uncontested summary appends for one position or the other. It can often be reasonably presupposed that this condition means that an informal rule exists. The best evidence of such a consensus is change in behavior on IBMPC. Desirable behaviors may be more apparent. Undesirable behaviors may subside. Particularly in the latter case, it may be possible to directly or indirectly observe informal enforcement actions, including private notes to apparent violators and modifications to appends that appear to violate the apparent consensus.
  2. a general appeal for tolerance. Many discussions on SENSITIV FORUM end with one or more uncontested appends that point out, as does this October 1984 excerpt from SENSITIV FORUM, that "if something in a forum is irritating, it could be that the writer is abrasive. It could also be that the reader is sensitive. Look at both possibilities before passing judgement."
  3. discussion ends in a draw through the intervention of a reviewer or a weekend. There are a number of interactions on SENSITIV FORUM which end without any apparent consensus. This kind of discussion most often ends when a weekend away from work (and conferencing) succeeds in cooling emotions. Sometimes more emphatic action is required, as was the case when a May, 1987 discussion over grammar and spelling attracted the following append by an IBMPC administrator:
    This is a plea for a truce, a de-escalation of the hostilities that seem to be close to bringing this forum and its participants to the brink of warfare. While I understand that there are at least two camps in this discussion, I do not see that these must be ARMED camps. How about less slings and arrows?

The functionality of meta-forums

The experience of meta-forums on IBMPC has been a positive one. Meta-discussions that could once be expected to destroy otherwise productive forums are now isolated to a separate place. The resulting discussion can be every bit as intense as it once was on regular forums, but the patient no longer dies (at least not without explicit action on the part of the IBMPC administrators). The meta-discussion now proceeds in parallel with whatever other discussion provoked it, and without disrupting the flow of subsequent appends.

With discussion isolated to a repository of such discussion, there is a much greater tendency to apply the lessons of prior meta-discussions. Hence one frequently sees back references to prior discussions or cross-references to discussions on other meta-forums, which may well provide a useful precedent for guiding discussion. Used in this way, the meta-forums act, in some sense, as a repository of IBMPC's informal rule-making process. Indeed, when one takes the trouble to look, one can frequently find the meta-forum discussion that led to a given informal rule.

The meta-forum is not, of course, the only means by which people meta-communicate on IBMPC. Meta-emotion icons are used to clarify the intent of messages. Private notes are used to apply peer pressure and clarify meanings. Feelings about what others have said will frequently (and often unintentionally) spill over into appends. It may well be that interaction on meta-forums represent only a small fraction of the meta-communication that occurs on and around IBMPC.

Still, the meta-forum represents a rather unique distillation of a key element of the normal process of human communication. The assertion that communication acts are governed by mutually negotiated informal rules is a common one in the literature of human communication. The existence of such informal rules is, however, usually argued from their effects rather than from observation of the process of their negotiation. The process of the meta-forum is the process of negotiating the rules of everyday interaction. It provides a unique means by which the IBMPC community can and does interact to determine its own informal rules of conduct.

Rules and Media

The overt negotiation of rules in written computer conferencing meta-discussion is probably fairly unusual among media. Most meta-communication is tacit, expressed through nonverbal give-and-take. The hazards of meta-communication ("murder by meta-discussion") in computer conferencing may be unique to the medium. Its combination of a face-to-face interaction style with a mass media audience creates a potential for clearly individual, yet rapidly accumulating, expression that may allow discussions to go out of control more easily. The structures of meta-communication (the meta-forum) once appeared to be unique to the IBMPC computer conferencing facility. Meta-forums can now be found on most of IBM's larger conferencing facilities, and the practice can also be observed on other company's in-house computer conferencing systems. Still, while overt meta-communication is hardly exceptional on computer conferencing facilities, meta-forums remain something of an exception on the computer conferencing landscape.

The process of rules formation that can be observed on almost any computer conferencing facility is hardly exceptional. The conduct of human interaction requires that we adapt our communication behavior to meet the needs of others. It is through the negotiation of rules, whether done tacitly or overtly, that we agree on what kinds of behavior are acceptable. The negotiation of informal rules on IBMPC, as observed in its meta-forums, may be more obvious than it is in other media, but the general processes by which they evolve are probably the same on all media, whether mass or interpersonal.