Aspects of the media themselves

I'll take it as a given that MUDs and other such conversational media will display unusual sociology; they are discussed in much greater detail both in my examination of muds themselves, and in my examination of how anonymity and pseudonymity tend to shape their sociology.

Instead, I'll take mail and its spinoff media as case studies. In particular, I'll look at

• hardcopy mail (AKA physical mail or snailmail)
• electronic mail
• Zephyr
• Usenet news

Let's take the first three media above. Why am I examining them? Because they illustrate that, even though they are essentially purely print media, subtle changes in their characteristics can lead to large sociological differences.

For example, it used to be, before telephones were widespread, that people used snailmail for informal, personal correspondence. With the rise of the telephone, that usage dropped, and most snailmail these days is commercial. Those who still use snailmail for letters use it mostly for invitations and other somewhat-format communications needs; actual, chatty, personal correspondence seems a vanishingly small percentage of the whole.

Yet with the rise of the net, the use of email for personal correspondence has seen a renaissance. Email doesn't require a postage stamp, finding a mailbox, envelopes, and so forth, and furthermore is implemented in a medium that make editing much easier (typewritten personal letters were a compromise form in snailmail---they had uncertain social approval and often stradled the boundaries between formal, commercial communication and informal, personal communication). It also supports a more cocktail-party style, many-to-many mode of communication in addition to its one-to-one, personal style.

It is instructive, however, to observe the social pressures that are acting in certain environments to formalize email in the same way that the telephone formalized snailmail. Take the MIT Athena environment, for example. This environment supports Zephyr, a text-based, instant messaging system with three major differences from email:

• by default, messages pop up on recipients' screens, rather than waiting passively to be read;
• by default, messages are ephemeral, and it is somewhat inconvenient to save them;
• if the target user is logged out, the sender will be warned before transmission (messages to an instance to which the user would otherwise be subscribed are simply dropped)

Yet among MIT undergraduates, who are the primary users of Zephyr, the apparent ephemerality of Zephyrgrams makes them the medium of choice for quick questions, invitations to find dinner, random chatter, and so forth---which used to be a primary discourse style in email at MIT, and which still is in many other places which are heavily email-dependent but do not run Zephyr. Given that Zephyr has co-opted such ephemeral chatter, messages sent via email are often viewed as formal and somewhat commercial or official, even when they are sent one-to-one and even when they are about similar subjects. Zephyr has colored the reception of email in the same way that the telephone has colored the reception of snailmail.

Small differences in media like that make all the difference. For example, consider both large-scale email lists and Usenet news. These are similar media; news was established to compensate for deficiencies in dialup UNIX servers which were not, at the time, connected to the ARPAnet (now the Internet), whereas mail was a part of the permanently-connected hosts on the ARPAnet almost from the start. Yet the broadcast nature of news (all news is flood-filled to nearby hosts, whereas mail follows explicit paths and is essentially nonbroadcast, even to lists) has exerted pressure to hierarchicalize the various newsgroups (since otherwise there are so many that chaos would reign), to come up with the ideas of Frequently-Asked-Questions (FAQ) lists so new readers could come up to speed quickly without spamming the entire newsgroup with familiar questions, and so forth. (This is necessarily a very fast and sketchy account of the difference in their sociology; I aim here for illustration, not completeness.)

Consider also how much different the net would look today if a simple technical detail, such as whether charges are flat-rate, dependent upon the bandwidth of the connection, or per-packet, dependent on amount of traffic. X25 links are generally charged per-packet; as a result, widespread, collaborative networking never caught on in Europe, where such connections are the norm, until the spread of the Internet, bandwidth-based charging scheme became common. Indeed, one could make a cogent argument that, if the Internet had charged per-packet from the very start, it would have had a crippling effect on the development of mailing lists and newsgroups, and such forms might possibly never have arisen.

Incidentally, One might argue that newsgroups, having originally grown out of the switched telephone network, wherein computers called nearby neighbords and transferred traffic that way, would have arisen even if per-packet changes on the main Internet existed. However, that analysis ignores three points:

1. the Usenet originated after the model of the Internet's (then the ARPAnet's) mailing lists existed as a proof-of-concept
2. most of the Usenet traffic was via local-call phone hops, which generally are flat-rate (because they do not charge for line time, and packets/second is roughly constant)
3. the bandwidth available for such local hops is sufficienyly limited compared to the long-haul bandwidth available on the main Internet (where the vast majority of Usenet traffic now flows) that, if the Usenet still had to depend on dial-up phone-by-phone hops, it would collapse under the current load imposed (and messages would still be taking a week or more to transit the diameter of the Usenet, instead of the typical several-hour transit time we see now)

Back to Contents