User:Mshonle/Around 2000 Words

We're a lot of things. I have hopes that our wiki textbook project will change the world. But I don't think we're wiki-people who do textbooks. Rather, we're textbook-people who use wikis.

We Have Unique Problems

I think that we have unique problems. Beyond the problems that face the Wikipedia we have additional challenges. For example, an article on the Wikipedia about Fidel Castro really only has one direction to take: give the facts, the dates, the places, the important people, and the important events. I don't mean to trivialize the effort required to make a high-quality article on Castro -- it takes a lot -- but I do want to show the challenge of writing a book.

For example, how should a book on algorithms be written? There is no canonical form of the "perfect" algorithm book to turn to. One way we write our books is to start out with a table of contents, reach consensus on it, and then restrict ourselves to serving those goals. Sometimes our goals change, but it's better for attracting contributors and for reaching a proud "first edition" to have a goal spelled out early on. The most valuable kind of goals are those that tend to encourage the most educational or instructive results.

Our project has finally matured to the point that we need to define the goal for Wikibooks itself. We are all asking ourselves, What really is a wikibook? Where do we draw the line, if at all?

Information is Good

Thinking deeply about this, I came upon a lot of unworkable ideas: Wikibooks should only have what's available in the library. Wikibooks should only have works where there are two like it in the library of congress. Wikibooks should only be for courses taught at accredited schools. In each case, they left out too much.

I know that some of you feel strongly that there is no such thing as leaving out "just enough": even a single exclusion would be too much. But what happens when we follow this value to its logical conclusion? Imagine a cookbook that contains every recipe anyone ever tried. It'd be terrible. After trying the first five recipes, you'd stop because the results were so poor. The fact is that most recipes aren't that good. I'd much rather have a cookbook with only 1,000 decent-to-outstanding recipes than a cookbook with 100,000 poor-to-outstanding recipes.

I think it can go without proof here that we all want to see true information do its magic by reaching free minds and helping people reach their full potential. But I also think life is short. I once started to study computer programming by reading some decent books, but then I followed up with some mediocre books. And I simply didn't know the difference. Someone more experienced then pointed me to some great books to read. I suppose I wasn't surprised that the books were 20 years old, because this individual was so good I didn't doubt his advice. In a large world, with limited time, quality matters.

Currently we filter out the less-than-quality books through our vote for deletion process. Although VFD is supposed to be a means to reach consensus rather than a raw vote, often it can be a haphazard process. I don't think it's really fair to contributors if we first tell them they can create any book they want if we then later end up deleting it. Similarly, I don't think it's fair to the other contributors (who do produce high-quality works) to have the reputation they earned for Wikibooks get diminished by association with works the community or the founders deem unsuitable. We know there is a problem with our policies somewhere.

The Wisdom of Crowds

One suggestion that gets a mention is to use a rating system. This idea is particularly strong with the cookbook. It might be convenient philosophically to follow such an approach. For example, we could do the same for books, and only let the high-quality books reach the front page. We could introduce a book recommendation system to increase the popularity of newer and lesser-known books.

I think for such a plan to work, Wikibooks will need to have grown a bit, and have a large enough "seed" of books to attract serious readers. But I see another problem with such a system. I don't doubt for a second the "wisdom of crowds." Now, sometimes crowds aren't perfect, but they've shown amazing results again and again. But group averages work best when each party has an actual stake in the outcome. For example, when you buy a stock, you are putting your own money on the line. Through the magic of markets, the price of a stock reflects (more or less) the currently known information about a company and its chance for success. Betting systems have been able to predict presidential elections with high, but not perfect, accuracy.

But I'm not here to defend that view, because I don't think a system of votes for popularity and recommendations will work well for our site. The recommendation system of Amazon works because it bases (some) of the information directly from what people have purchased. If someone wanted to play with the ratings and game the system it will end up being very expensive for them. On the other hand, we are all aware of sock puppeting. In the presence of such disruptions the results are skewed enough to be meaningless.

A Proposed Policy

In reviewing what Jimbo has said, how VFDs turn out, and considering the opinions of all of the voices involved in our project, I have, with the help of others, enumerated a policy I think that is appropriate for Wikibooks. This policy reflects my opinion that I'd rather have a Wikibooks site with 1,000 decent to great books than a Wikibooks site with 10,000 terrible/offensive to great books. Actually, let me restate that. I'd rather have a Wikibooks site with 100,000 great books!

But already I've triggered some subjective terms. Offensive to whom? To whose standards must we conform? While I believe we could have great results with a liaise-faire, anything-goes, all-legal-books-are-allowed approach, I doubt any results will be realized until there's a rating system based on people who are putting something of real value at stake.

Until then, we need real standards. This includes both standards of quality and appropriateness. It also includes providing a sharper direction for the project. Remember that the Castro article on the Wikipedia already has the "canonical encyclopedic entry" standard to head to. We need our own direction, because modules do not turn themselves into complete, coherent books by the wisdom of crowds. The only time I've seen a book become complete, coherent, consistent, and worth reading is when a small group of individuals coordinated heavily and took commanding roles. (I am, however, sure that the proof reading, rewording, spellchecking and other improvements to an already completed work need no coordination.)

My thoughts on the standards of quality, for example, requiring books to have a direction, and for all modules to belong to an on-topic book, seem to be agreeable. My thoughts on where to draw the line are the most controversial.

For example, I think it would be wise to follow the letter and spirit of US law. Actually, in this regard, we don't have a choice. But even more controversial than that is if we should adapt rules similar to the Google AdSense prohibitions.

Such an adaptation would not be unheard of. Content on Wikicities already has such restrictions. It would also create the option for people to host our books on their own sites. (Indeed, they will profit from the fruits of our labors, but they will serve to provide a mirror when our site is down, they will do any promotion themselves, and they'll likely include links back to Wikibooks-- This all means that we can reach more people, which is the reason we write.)

I think most surprisingly is that such a list of exclusions, which is likely based on what makes advertisers feel the most comfortable, has very little, if any, limitations on the role of being an instructional resource.

I don't think classrooms are useless for the real world either. In particular, with the "information helps the victims" argument, there are responsible ways to enable people to take precautions, and then there are reckless ways. By sticking to the responsible side we improve our reputation, and we can become trusted guides. In some ways the exclusions serve only to make a speedy delete out of what currently needs to go through VFD. Encoding this wisdom we've gained into rules makes for a more consistent site.

Also, I have doubts that any book written directly for these dark gray areas can evolve into useful resources. The more base books, which were guides to breaking the law and intimidating or harming others, were canyons away from ever becoming an honest book to aid crime prevention. I don't think honest people would even desire to do what's necessary to turn those books into good guides.

I believe that if there is something worth knowing that it can be taught responsibly and in a proper instructional context. I provide the following reasons why I think Wikibooks should avoid the dark gray areas, and allow any attempts at creating them to become speedily deleted:

• By being closer to Wikicities's policy we can more easily convert works to and from that site to our site
• By following the AdSense rules we are creating valuable economic incentives for others to create mirrors of our site
• By not advocating the darkest areas we will have less danger of tort action; it's not fair to those writing mathematics books that the site would be shut down based solely on one misguided book
• By adhering closer to community standards, we are leaving open more doors, so that the site will be able to be accessed from more schools, more public libraries, and more organizations than perhaps otherwise
• By respecting community standards, like those practiced at universities, we are more welcoming to expert contributors, who might only spend large amounts of time at sites that they could list on their CVs without embarrassment
• By constraining the way we frame and present material, rather than restricting the true knowledge beneath, we are guided more toward instructional exposition and guided closer to true books for learning

I know that it pains many for this project to "censor" itself (which implies, ultimately, the censorship of some contributor). But the gains from a project taking editorial guidance should not be confused with the evils of government censorship and oppression. We ultimately want to teach others what we have learned. The true, worthy lessons that can be gained from dark gray areas can be taught without the dark gray baggage itself. The smartest people are able to connect the dots for themselves. That is the nature of knowledge: it can be used for good or evil. We will never get around the fact that some people could use what we tech them for bad purposes.

But instead of teaching people how to make bongs, we should teach general craft making. (You can figure out the final step yourself.)

Instead of teaching people how to make explosives, we should teach chemistry. (And after really understanding chemical reactions, instead of just following directions, you'll know the true dangers enough to make an informed decision for yourself.)

Instead of teaching people just the human capacity of hate, we should teach the humanities. (You can then understand both the pathetic and the triumphant nature of mankind that way.)

As educators we must take the ethical path, make the hard decisions, and give serious thought to important legal or economic considerations. But most importantly: We must be a good example for others.