The problem is more than just writeups that are wrong, but also that a lot of factual noding here is not adequately supported.

The other day I had an interesting exchange with another noder about a certain writeup of his. It's a great writeup, with lots of interesting ideas about linguistics that I hadn't seen before. I had a problem with it though: I don't know if I should believe all of those ideas.

I don't know if I should or can believe him because he didn't give any sign of where I could go to verify what he wrote. I would have liked to see the name of a linguistics text or two that might support his premises. I voted it up, but then i messaged him and asked him about his sources. We went back and forth and it ended up that he basically proclaimed himself an expert, and claimed he had talked to other actual experts in the field, but that he had "no idea what the right texts are". This quite frankly baffles me; How do you become an expert, especially in an academic field like linguistics, without knowing some of the literature of that field? (Is it transcendental knowledge?)

I guess I'm a pretty academic person, or have been lately, reading lots of books where I need two bookmarks, one for where I am in the main text, and one for the footnotes (which nowadays are mostly endnotes, but I still call them footnotes). This despite the fact that I've been done with graduate school for 5 years. I still consider there to be a lot of things I want to learn - which is probably one reason I'm interested in E2.

So that's my bias. I'm interested in knowledge, not just information. I'm interested in verifiable, proven or provable, supportable, cross-referenced, credible, data. Not hearsay, not hypotext, but real, hard, trustable stuff. That's not all that Everything is about, of course (!), and everyone is entitled to their opinion too, of course - but even when you have an opinion, it helps to note why the hell the reader should care: Why should your opinion count?

And I need more of an answer than "I'm an Editor/God/Level 7/etc". Just because you've logged a ton of hours and bytes on this website doesn't cut it in my "credibility scale". And if your homenode says you're specialty is X, that doesn't cut it either. I want to know what books you've read, and/or how many years you've studied, who you studied with, what papers in what journals you've seen. I want to know what department of what university you're faculty on (not that this is neccesary here on E2 - I'm just saying that if you have some kind of commonly-respected credentials in the field, it doesn't hurt to mention them).

If E2 is ever to appeal to more than just a few otaku, it needs to have more attention paid to credibility. This is even more important if the "Node for the Ages" idea is real and not just lip service. In 60 years when many of these writers and the people they know are dead, if E2 is still around readers will want to know who they were and why they should be believed. Probably no one will remember who I am. But maybe they'll remember people and sources that I cite in my writeups.

I can anticipate that an answer to this will be that E2 is "self-contained" and therefore has no need of outside references. This is a pipe dream. Maybe someday it will really include everything, but probably not, even if there were no copyright. However, even if all information was contained in E2, you need something else: all people. E2 would need all of humanity, or at least all of literate humanity, to be members, and to have complete faith in the E2 reputation system and "caste structure". And face it, that ain't gonna happen (at least with the present system).

Let's get real: people who are smart and have good ideas and can support those ideas should be paid attention to, whether or not they're level 1 or level 10. If Larry Wall started noding about perl or Noam Chomsky about linguistics, people would damn well respect them, I reckon, even if they are only level 1. And let's face it, Larry or Noam would never have time to be more than level 2 or so. They've got better things to do, god bless 'em. (Paradoxically, if all you ever did was node, you wouldn't be much use to Everything, would you?). Realistically, Noam and Larry and millions of other smart people will never even create an account, but luckily, others of us here are able to input some of the valuable knowledge they have created in this world, and, hopefully, attribute it to them.

okay, this writeup has now been cooled 6 times. stop it, please. it's really not that great. in fact, stop even voting for it. thanx.

Let's for a second look at other factual (or "factual") works that people rely on: dictionaries and encyclopedias. I'm not only going to show that E2 is almost identical in its form to those literary forms, but that if you have a problem with this in a hypermedium, the answer is a radical overhaul of our metadata management.

Consider the Encyclopedia: a collection of short articles summarising topics. Organised by title in alphabetic order, and searched by title. Uses "see also"'s to reference other articles. E2 nodes are searched by title, but often found using links within other articles. E2 also associates author metadata, and original date metadata with articles. Votes are also a form of metadata. As authors can be traced and called-up on the content of their nodes, E2 WU should be more accurate than the anonymous write-ups in an encyclopedia. E2 nodes also have a guide as to how old (and hence out-of-date) they are, which encyclopedia entries lack. As such, E2 wins a credibility-off with most encyclopedias, excepting the Encyclopedia Britannica, which has the same metadata (excluding votes).

The one difference with the Britannica is that Britannica articles possess metadata about the source of the information. This is delivered in the easisest form to navigate, alphabeitcally ordered, keyed, citations. E2 is a computerised system, however: there is no reason to simply use the forms that worked best in books, because we have more alternatives. If you want to know more about your articles, and you find most encyclopedias to be unsatisfactory (As you must if you find e2 unsatisfactory), then perhaps the solution is to make better use of the medium.

A starting point (Notwithstanding my position above) is to computerise the bibliography. This would allow hyperlinks to the full citation, to summaries of the books, to the authors. One could search articles by their citations. We could then add the opportunity for authors to provide structured metadata about each kind of entry: Person entries could have names, birth dates, lists of important dates. Every single entry could have keywords entered. Places could have structured names. The possibilities are endless. Don't settle for the old and the dusty, when we can have the old and the chrome-plated. Thank you for listening.

This is really part of a larger problem: the Internet credibility problem. In a nutshell, how can we trust people online with important information? How do we put our own knowledge out into the cyberether for people to see and use? With the Internet and associated technologies still in their infancies, we really don’t understand how to handle these issues yet.

Take e-mail. George Carlin said that e-mail is for communicating with people you really don’t want to talk to, and I think he’s right. E-mail is in a weird position right now, competing not only with traditional means of communication (especially postal mail and the telephone, although sneaker mail is rapidly being replaced in offices), but also with other emerging technologies: instant messaging, videoconferencing, chat rooms, and forums, among many others. And we’re not quite sure what the role of e-mail is or should be.

Think back to grade school: you were taught how to write a business letter, how to write a friendly letter, and maybe some other kinds of letters besides. These weren’t taught as a set of conventions that had slowly gelled — they were taught as absolute rules. They may as well have been written into some Constitution or brought down from Mount Sinai for the masses to absorb. But the point is that everybody else uses these conventions. I know that if I were to write a business letter using the model taught to me when I was but a lad, it will be indistinguishable from other business letters written by professionals who were writing them from experience, and not from the Rules.

But e-mail we don’t know how to handle. Do you put in a salutation, or don’t you? I usually don’t, but I’ve been told that when I’m in communication with somebody who does do this, I should as well, to put them at ease. How formal should the language be? Is this the same as a normal letter, paying attention to phrasing and structure? Or is it made less formal by the delivery speed and the fact that one is free to write as many e-mails to as many people as one wants? Do you put your name at the end of the e-mail, despite the fact that the message was probably already tagged in six different places with your name? And I won’t even get into the subject line — that bugbear of anybody who has to sort through a lot of mail.

We don’t even have this problem with most other forms of computer-based communication: I would argue that there are accepted styles of communication in place already for many media. Instant messaging, for example, imitates conversational style. I’ve found that people who speak in slang and fragments are preëminently the ones who IM always in lowercase with that annoying AOL-er habit of contraction, whereas people who speak in full sentences full of subclauses and implied semicolons tend to write instant messages the same way. There’s also a business style, for those working stiffs (un)lucky enough to have Microsoft Messenger integrated into their workplaces.

Forums we know how to deal with; they’re a slightly delayed version of Instant Messenger, with a stronger focus on argument. Much of the same writing style persists, but with more attention payed to paragraph structure, spelling, and thought organization. Cellular text messages also are a solved problem: they’re the modern equivalent of the telegram. The objective isn’t to send a long, prosaic message; instead, people shorten the message, using as much implied language, slang, and the like possible to aid rapid, informative communication.

This problem of inconsistency isn’t a product of our times and technologies. Throughout human history, there’s been an adjustment period to new forms of communication. I posit that one of the largest concerns, if not the single largest concern, with new communications is trust. Can you trust the person who is trying to communicate with you? Can you trust that your information will make it to the person you're trying to communicate with? And how can we use what we know to make the new system trustworthy?

Think about why we have the phrase “don’t shoot the messenger” — there was a long time in communication between enemy states when it was considered perfectly fine to execute a messenger who carried bad news. It took time for people to decide that the messenger was just the passive carrier of information. In large part, this because that a messenger could be compromised: there was nothing stopping somebody from replacing the messenger partway along the route with one carrying a different message. Of course, an effective deterrent to would-be messenger compromisers would be to kill them. And until an effective solution to the problem could be found, it would be very hard to convince a leader to stop killing messengers (and seemingly encourage them to be corrupted again). This is part of the origin of signets and seals: if only the king had the ability to put his wax seal on the message, the message could be trusted to be from the person whom it claimed.

Similarly, the first times that dictionaries, encyclopedias, and the like were compiled, they were riddled with errors. It took quite a long time before dictionaries got to the point where they could be considered the final authority on language. Have you ever thought about the fact that a dictionary is compiled by normal people, researching words, who can (and do) make mistakes? If you do, it’s most likely a whole lot less than people thought about it several hundred years ago, when they were a new idea.

Similarly, our beloved Everything, as well as the Wikis and the H2G2s and all of the other burgeoning online references are not a form of communication that we understand yet. We’re trying to impose the lessons that we learned from older analogues onto their newborn relatives, and it only works part of the time. What is particularly interesting is that the structure of the problems isn’t new, it’s their content.

The problem of accountability arises again. How do you know when you read online content that it was written by an expert? It could have been written by somebody who was misinformed, or who got one critical fact wrong because their research wasn’t solid enough, or, in the worst case, is deliberately wrong in order to spread a different message. This problem arises far more commonly in the Wikipedia model, where accountability fluctuates. Just because an article is trustworthy today does not make it so tomorrow, and there aren’t enough editors in the world to check all of the content.

At Everything, we have a different sort of accountability problem. Although we don’t have to deal with the fact that information is supplied by single, uniform source (try all you want to convince me, but a list of IP addresses still won’t be an author), we do have to deal with the fact that different noders have different backgrounds, sources, and writing styles. Even if you’re reading a piece that’s well-researched and cross-referenced to scholarly literature, you may misinterpret the writing if it’s in a different style than what you’re used to. Print encyclopedias learned long ago to enforce a uniform research and writing style to ensure that a reader could approach any two articles in the text with the same attitude and reading methods; this seems to fly in the face of what we’re trying to accomplish at Everything.

In fact, citations don't necessarily add anything to a text. All that they say is that some other source agrees with the citer. We've all seen citations of documents that take the originals completely out of context, warp their meaning, misinterpret them. We've also all seen citations of disreputable works. The only thing that citations add is a gloss of authority, and even in the best case all it tells you is that some substantial body of people agree on an issue — not that they're right.

We’ve made a step in the right direction by linking content to specific authors. Even if you don’t universally trust Everything as a fact source (and if you do, you should go read one of the nodes dealing with the accountability of drug-related noding), you can still have a mental short list of dependable writers. You may know for a fact that a certain noder is an expert in her subject and a respected authority in the field, which gives you a great deal more confidence in her writing. Conversely, when you read a node from a noder unknown to you, it’s feasible and practical to research him and find out what else he’s written, and how dependable it is.

Ultimately, the problem with these emerging encyclopedias is that technology as it is now makes it very easy to create content, but very hard to check it. You could spend hours going through IP logs on Wikipedia to make sure that nobody with a vested interest in disinformation had a hand in the article, but in the time that you’re doing that, it can be updated and changed many times over (not to mention the fact that this will be happening with many other articles simultaneously). Similarly, although the New Everything will have a certain amount of review before writeups go live, and H2G2 already does have a primitive form of editorial review, there is a great deal of misinformed content that can slip through the net.

These problems will sort themselves out. If we never find a way to ensure that community-supported references are trustworthy, they will become a second-class information source. There will always be a use for a trustworthy, fact-checked encyclopedia, even if it isn’t made available in print any more. Researchers and their readers will learn not to trust certain fact-checking styles, just as you can now dismiss a lot of the fluff pieces pushed through the printing industry by vicious lying pundits just by flipping through their sparse, untrustworthy bibliographies.

As I said at the beginning, we’re dealing with a new sort of communication, still in its infancy. It has a long way to go before we can fully trust it, but it is good to know that as long as it is being made, people are trying to make it work.


Thanks to XWiz for downvoting for a good reason.

Log in or registerto write something here or to contact authors.