Or the water cooler, to be more precise. You know, the thing people go to when they want to take a break from work and talk to someone to clear their head.

The evolution of private-public speaking and weblogging as a practice is a fascinating topic. Since it’s so fascinating, I’m sure, there’s a lot of good research on it, which I can’t be bothered to track down right now. Instead, I’ll say a few things about it, much of them from memory while looking to the Wikipedia with one eye, in a feeble attempt to write down a lecture that was supposed to be part of a course I never actually got to write.1 This story is neither finished nor sufficiently detailed, but it should give some sense of perspective. I’ll start with quoting this instead of an epigraph, because it’s such an apt thing to say:

…writing in public is a pain in the ass, especially in this modern age of the Internet, where anything you say can and will be used against you, by someone, at some point in the infinite future. After all, everything you say online will be available somewhere, some place, online, forever.

(from Shit my logs say)

In actual practice, it’s quite far from forever, though unless you plan specifically for this, the length of time things you wrote on the Internet may last inevitably turns out to be longer than you would like. It’s interesting how every epoch in the development of interpersonal public communication breeds the seeds of the next one, but the way it happens is different every time, and they all revolve around permanence and ease of access. But let’s start with the beginning…

The Ancient World

With these computers, our ancestors went against dinosaurs to toss ballistic missiles at them… ahem.

The practice of netiquette predates modern blogging and social network etiquette norms by so much, that by the time they emerged as a communication medium, it had to be reinvented. But what’s particularly interesting is that it first emerged on Usenet, which started in 1979 and did not, initially, depend on TCP/IP, which was introduced in 1974 – while the original ARPANET began in 1963. Apparently, ARPANET was seen as a “government business only” network, where non-work concerns were undesired, while Usenet was a grassroots effort – so grassroots that it didn’t even have such a thing as a stable network address.

It is important to notice a few things about Usenet:

  • A Usenet message, by design, is little different from a email message broadcast to a special interest mailing list, and mailing lists predate pretty much everything else, including both ARPANET and even email itself. A “mail merge” operation, that is, plastering addresses on a batch of identical xeroxed letters, was a staple of business practice since pretty much forever.
  • All members of Usenet were already a part of a certain social strata and established intermeshed social groups – they were members of the scientific community, both students and scientists alike, if only because no others had access to computers in the first place — starting with computer scientists, because they had more computers to go around. As I already mentioned a few times, in ‘softer’ sciences, scientific discussion is impractical to differentiate from daily discourse on all topics,2 while in ‘harder’ sciences, everyday discussion at the very least facilitates the spread of scientific ideas, which the whole thing really is about.3 They needed a medium for everyday social communication much more than they needed a medium for everyday business communication – they already had that.
  • There was a sharp distinction between “what gets written”, “what gets said aloud” and “what gets said before an assembly”, and all three bases were covered – what gets written is a paper in a journal, what gets said aloud is what you say around the water cooler, and what gets said before an assembly is what you say in a lecture, which is different from what you say in a textbook you write a year later, sometimes substantially. Usenet took the place of the water cooler, even though it was, in fact, written.

The “government-business-only” ARPANET was soon supplanted for these purposes by CSNET (1981), NSFNET (1985) and other major general purpose network backbones which catered primarily to universities, where Usenet quickly migrated, and soon the coverage among the scientific community of USA was pretty much universal.

This wasn’t the first world saw of interpersonal communication which shifted away the time and space constraints that make it inevitably private,4 but it certainly was the most extensive to date, and notice that just like the concept of a scientific journal grew out of exchange of private letters between rationally minded individuals interested in pursuit of truth and acceptance of their peers, so did the Usenet culture, which went on to discuss… well, everything.

One other particular thing of note about the Usenet is that there was no registration or access control necessary up until it was supplanted in the role by newer mediums, which had to invent at least some barriers to protect against spam. Whatever access control needed to happen has already happened — you got an email address assigned by the sysadmin of your scientific facility, and that identified you sufficiently, the same sysadmin controlled access and only allowed in people whom he could justify spending the limited computing resources on. It really was scientists talking at the water cooler, because it was their water cooler and non-scientists only got in by special dispensation.

While Usenet is still in widespread use, its usage patterns differ significantly from its golden days, and now it is more commonly seen as a means of binary file transport than anything else. The start of its decline roughly corresponds to the rise of HTTP, that is, the internet-as-most-people-think-of-it, and the change in general social makeup of the Internet.

The Atomic Age5

For a while, the mode of retransmission adopted by Usenet, which is an ad-hoc propagation tree, was basically the only way to go, because the Net still contained a lot of impermanent dialup links even between major population centers. A client-server model was simply impractical where the server might not even be up at this particular time of day. By the time Tim Berners-Lee went public with his WWW project in 1991, however, the long range backbone networks were already well-established. Notice that it was still mostly communication “by scientists, for scientists”, and WWW was meant to provide a text with links — i.e. supplant the paper in “what gets written”, because people do not actually talk with hyperlinks, or at least, did not at the time. Berners-Lee himself called it a ‘database’ rather than anything else, probably for precisely that reason. The Gopher protocol, which was published just months before Berners-Lee’s proposal, was in fact expressly organised as a database for documents.6

What made WWW win everyone over eventually, though, was the introduction of graphical browsing and rich text in 1993 with Mosaic and the subsequent commercialisation. As people who no longer were students wished to keep their email and Usenet access, corporations began offering it, and the era of America Online, CompuServe, and other similar services produced a mass swelling of networked population, both from people who were previously exposed to the culture and not. 1993 is actually the official start of Eternal September, the colloquial term for the never-ending onslaught of new users the existing culture was never able to completely indoctrinate.7

With the yearly cycles of new students appearing, hardly anyone kept track of students leaving. A PhD student, however, will spend at most about ten years in education, and it is no coincidence that the shift in preferred interpersonal communication medium also corresponds to a shift in generations.

There were two important things that made a webpage radically different from a Usenet posting:

  • While anyone could, in theory, subscribe to any newsgroup, acquiring its public archives, if they even existed, was a much more complicated task, so what has been said sufficiently long ago was definitely water under the bridge, whatever lore there was could only be in memory.8 With a webpage, the author of a webpage was in more or less total control of whether something they published remains publicly available,9 and if they, for whatever reason, lost their web space10 the data would usually go away with them.11
  • Up until search engines got up and running, and for an extensive period afterwards, the most practical way to locate a web page was to come in from a link posted on another webpage,12 while Usenet postings by design implied a specific interest group with a well-defined topic. This, instead of diluting the water cooler environment, which the public availability of every webpage everywhere would imply, actually strengthened it — it was simply not limited to scientists anymore. Notice also the term “homepage”. It derives from the browser’s ‘home’ button, and at least in the early days of the Web, it was literally a publicly accessible page which the user of a particular browser would use to keep frequently used links — literally, the social tapestry made manifest in its most abstract and chaotic form.13

The earliest traditions of blogging grow smoothly out of the concept of a webpage — records of changes to the content arranged in a reverse chronological order closer to the top is certainly an ancient idea that dates to this period. The earliest blogs are supposed to be dated to at least 1994 and undoubtedly derive from someone simply being meticulous about updating their homepage.14

What made blogs a new beast and made them a popular form was not the method of publication, though, but the form of writing itself, which made them a distinct movement — it was even described with the word ‘escribitionism’. It is an unusual hybrid between “what gets written” and “what gets said aloud” — the mode of writing used was typically epistolary rather than journalistic or straight out conversational, the kind that gets used in a written personal diary. Notice, once again, that it’s still a personal diary, not a pulpit. Publishing a personal diary ‘after the fact’ is an accepted practice with much historic precedent, just like writing a personal diary is a practice that dates at least to II century AD. Publishing it as it is written wasn’t, up until the WWW era.15

This is how the blogging movement started, but this was not what it grew into.

The Social Network age

What made the emergence of weblogging as a defining feature of Internet communication possible is the establishment of blogging-platform-as-a-service, which in turn was made possible by advances in web server programming, specifically all the work that went into permitting large-scale dynamically generated database driven web services during the 90s, to make previously existing databases more accessible. The actual creation of platforms as a service was motivated by aficionados of blogging as an online diary, precisely to bring the capability to people otherwise not technically skilled, or not willing to manage servers.16 These platforms were also the first widespread means of actually producing a reader comment which instantly turned every blog posting into a random street gathering.

LiveJournal is particularly notable in this regard both because it was one of the first17 such platforms, and at the same time, the first social networking software in the modern sense, formalizing the concept of a social link and picking the worst word possible for it — “friend”. At the start of the movement, they might as well have been, as the idea grew out of the basic concept of mutual linking between fellow writers of online diaries, which typically denoted a real social relationship. In LiveJournal, this even meant a certain level of compartmentalisation, as you could limit a particular post to a specific group of ‘friends’. However, it implied reciprocity, while diarists wrote explicitly for nobody in particular, unlike, say, journalists, or even members of a certain social circle, which wrote with a specific audience in mind.

Intermixed in the newly created environment, three groups of writers emerged:

  • Actual diarists writing for nobody in particular and using this as a platform for social communication in the what gets written form.
  • People not skilled or interested sufficiently in writing, which used it for what gets said form of social communication, mostly restricting themselves to comment and short postings.
  • Actual journalists, seizing the free-as-in-beer press to make it free-as-in-speech pulpit for an audience of persons undetermined.

All three would switch both writing style and goals left and right, so they can hardly be called very distinct, at that.

LiveJournal also gets the honour of being the biggest blogging platform in Russia, which skipped over most of the WWW era by having an established FidoNet culture, with its numerous similarities to Usenet and many unique quirks, transition directly into blogging-platform-as-a-service stage. For decades, the word ‘blog’ was not in common use in Russia because every ‘blog’ was called a ‘LiveJournal’.18

The eventual introduction of Facebook in 2004 and Twitter in 2006 gave rise to social network platform proper. In the earlier days of blogging-platform-as-a-service, it has become fashionable to segregate postings into postings ‘proper’ and ‘aside’ postings, of which, the proper postings followed a diary style, while ‘aside’ postings followed the original ancient usage of a homepage, which is adding a new hyperlink to it, with a very short comment on what it is. Twitter postings are exactly the latter,19 while Facebook is neither here nor there,20 but is still geared towards quick interpersonal communication in short bursts. Twitter might not be particularly responsible for the social networking aspect, but it is important as the epitome of the ‘short bursts’ ideology, while Facebook is now the archetypal ‘social network’, explicitly meant to connect you to people you otherwise know on an interpersonal level.

Every trouble, just like everything that is beneficial in this system, results from when interpersonal becomes public.

State of the World-P

Few things of particular note have been invented since Twitter.21 While there is a general trend to decentralisation and more open standards (OpenID, OAuth, the movement for data export, etc) and a few nudges towards outright federation,22 this trend is largely motivated by the constant tug of war between the pressure of powerful groups and the desire of the rest of the world to keep their water cooler and have the pulpit too.

In the beginning of the trend, the conflict did not exist, because the Net was limited to a population which is socially permitted to be more free and liberal than the rest of the society — primarily academics — who instinctively treated censorship as a malfunction and routed around it. The mechanisms they left as their legacy to the whole thing still do, whether anyone likes it or not. However, as generations replaced each other and rebuilt it using new technologies, what was previously, — and in essence and style still is! — private speech, or rather, the kind of speech one employs while talking with people they know personally on the street in full view of public, not caring who listens in, conversation meant for specific interested parties or nobody in particular achieved the qualities of permanence and accessibility. It happened neither quickly nor obviously, but that’s where we are now.

While permanence is not at all new to human history, accessibility is. What the author cited in the very beginning of this lecture complains about is not so much the public availability and the persistence of something said on the Internet — a postcard has both of these properties and they have been in use since at least 1840, and Usenet had both well before modern Internet came about — but the fact that it can be found by keyword in a search engine, without which, neither would matter anywhere as much.

That this is a particular worry, though, is because it’s the other end of using online communication as a pulpit — once you have something of social consequence to say, whichever is the way you say it, as soon as someone sufficiently powerful disagrees with it they will prefer to shut you up, and if they can find you proactively, rather than after what you said has done its damage, so much the better for them. At the same time, you rely on those same search engines to connect you with like-minded people, and just like in the Usenet of ancient times, there are no access control barriers on these things, so this accessibility is a double-edged sword that cannot be thrown away.

The obvious answer to this has been the wide practice of anonymity, which gives the speaker the option of social bankruptcy. This, incidentally, is also the most obvious means the socially weak have to protect themselves. Unfortunately, legislation that punishes ‘cyberbullying’ on the Internet, or seeks to protect people against defamation, also commonly comes as a package deal with legislation that restricts anonymity.23 No law can possibly distinguish between different types of social discourse, because it’s not actually differentiated in any king of hard fashion, anyway…

It’s not at all clear for me which way is it going from here, but it is quite obvious that just as the existing social structure is incapable of handling mass interpersonal communication spilling out on that scale, mass interpersonal communication is not something that can simply be forgotten as a fad — it’s already being taken for granted. Both the social structures and the structures of the Web will have to change to produce something new.

In the meantime, the Internet is leaking.

In the meantime, the Internet is **leaking.**

  1. …and with the way the country’s going, am not likely to get a chance to in the foreseeable future, because it’s quite likely we won’t have an Internet around here in the foreseeable future. Oh well. ↩︎

  2. In sociology in particular, you basically can’t tell if you’re talking about a daily concern or discussing a theory, you might be doing both without even noticing. It can get worse with philosophy. ↩︎

  3. And don’t tell me physicists don’t get drunk at their conferences because I won’t believe it. ↩︎

  4. The first was paper letters. I imagine that when lifelogging becomes readily available we’re in for the next shitstorm. ↩︎

  5. Well, CERN, so I suppose it should be The Boson Age or somesuch, but it makes no difference here. ↩︎

  6. The important difference was that while Gopher mimics a file system, and works very much like a file system, HTML was, from the outset, envisioned as hypertext. ↩︎

  7. Notice once again that previously, a September was an established phenomenon related to indoctrinating new students joining the student body in September, and only with the introduction of non-student population on a regular basis it became an Eternal phenomenon. ↩︎

  8. Notice the similarity to a forum with a broken search function. Which is pretty much every forum I ever saw. ↩︎

  9. Unless someone took the trouble to archive it, which few people bothered to do. Even fewer bothered to publicise the fact that they did. ↩︎

  10. Which, until the advent of Geocities and other major free web hosting projects, was typically only assigned per-user in the same fashion as email addresses were ten years prior, i.e. from a scientific institution to its member. ↩︎

  11. Which, I think, is the reason for the establishment of Wayback Machine in the first place… ↩︎

  12. Which is what prompted the comparison to ‘channel surfing’ when watching TV and produced a now extinct neologism ‘web surfing’. ↩︎

  13. Earliest attempts at web mapping as a graph also date to no later than 1995. ↩︎

  14. It’s worth noting that the general decline of FidoNet in the US, typically dated to 1995, coincides with that date, as declining costs of local dialup Net access caused the greater Internet to gobble up existing localised BBS communities. ↩︎

  15. And neither was commenting on what was written, which was not a thing until at least Open Diary. ↩︎

  16. For some bizarre reason, the concept of a static site generator appears to be much more recent — (MovableType, released in 2001, was both a database-driven blogging engine and a static site generator, but for quite a while it was the only one, and embeddable comment services to make the whole thing actually effective only became practical around 2006, though they existed earlier) even though the web was dominated by perfectly static sites for much of its history. The only hypothesis I have to explain that is that modern static site generators typically rely on human-readable markup syntax, formalizing which is an idea that did not crystallise until at least 2003 — while before, people used to write straight HTML and think nothing of it. Well, there was no CSS in the early days to muck it up… ↩︎

  17. But surprisingly not the first — that honour belongs to Open Diary which launched just a few months earlier and is no longer with us. ↩︎

  18. It’s not written anywhere, but as far as I was able to find out, the particular reason is that the initial Russian-speaking users of LiveJournal were highly literate and engagingly writing Russian Jews in the LiveJournal localisation department, which saw to it that windows-1251 encoding was supported early on, when the rest of the world pretended Russians write in iso-8859-5. They served as the seed population and ensured that Russian liberals still congregate around LJ. ↩︎

  19. What surprised me when looking this up was that Twitter is not directly descended from them — the initial proposal for Twitter revolved around SMS sent to a group, to which Twitter was tied for most of its history. Nevertheless, this is the niche it ended up occupying, and many a standalone blogging engine includes Twitter support for precisely that purpose. ↩︎

  20. It’s not particularly suitable for a long text like say, the one you’re reading now. ↩︎

  21. Youtube, while a truly major thing, is mostly outside the scope of this text as it has as little as possible to do with writing. It’s also been around since 2005. Mind you, most of these things are only really visible as epoch-changing in retrospect, so something else might slowly be developing without us taking conscious notice. ↩︎

  22. Most notably Diaspora, but there are other attempts. I just wish we would switch to IPv6 sooner, because lack of address space is one of the things that makes a private server much more expensive than it needs to be. ↩︎

  23. There is no evil in the world that cannot be justified by protecting children, and it’s been getting progressively worse for the past ten years… ↩︎