?

Log in

Previous Entry | Next Entry

Sep. 20th, 2010

The following is a draft written for Less Wrong, which I decided not to post there. But someone asked after it, so I pasted it out here (because it was a convenient richtext editor). Unless you happen to be a veteran of that site, or just happen to be interested in the problem of integrating newbies to an intellectual community with a lot of preexisting shared culture, it is probably of no interest to you whatsoever.


In response to AnnaSalamon's call for feedback on a recent post, I commented that I didn't feel qualified to vote or participate because I didn't understand the post well enough. MichaelVassar very reasonably asked:

Have you read the sequences yet? If not, can you suggest a good way to encourage people who haven't yet done so to do so?

The comment thread that followed contains a bunch of good discussion, much of which came after I wrote this draft, so I recommend following that link if you're interested in the problem. What follows is an expanded version of my brief answer, as well as a more detailed description of my proposed solution.

I have read some of the sequences, but not all; when I examined my reasons, I arrived at these three.


1) The minimum suggested reading is not just long, it's deceptively long.

This is because it links to sequence pages, which then contain many links to posts, so an initial "Oh, it's only a few" became "... times eight, oh dear" when I clicked on one of the links. I'm unusual in that I enjoy reading posted signs, filling out forms, and following instructions, and I balked at this. Even a single long list of all the posts considered to belong to "core" sequences would have been preferable, because I could have seen up front what it contained.

As to the quantity itself, there's a paradox here: You're asking me to be so interested in the subjects of Less Wrong that I'll want to do a whole lot of reading about them, before I can participate in the discussions that have piqued my interest. If Less Wrong wants to attract people who are new to rationalism et al, allow them to nurture their curiosity before gorging them on background information. If it doesn't, why require the sequences?

2) Many of the sequences are either not interesting to me, or are presented in ways that make them appear not to be.

For example, the page on the Reductionism sequence begins:

How to take reality apart into pieces... and live in that universe, where we have always lived, without feeling disappointed about the fact that complicated things are made of simpler things.

Disappointing?! It's awesome! It's one of the most fundamentally cool and interesting things about this universe! So, wait, do I need to read this?

It's possible that after reading more about "taking reality apart," I might see why one would be disappointed, but you can't rely on me having read the sequence when you're trying to get me to read the sequence. A more helpful introduction would have given a summary of the idea of reductionism--say, a definition and an example of where it comes up. That would have made it immediately clear whether it would be useful for me to read or not. (If you're wondering why the introduction needs to convince me it's useful when I've already been told to read the sequences, I ask what you would think of an alleged rationalist who blindly accepts someone else's rating of utility. More on that in the next point.)

There are also subjects I just happen not to be interested in, but since I presumably wouldn't be participating in the discussions about them, I feel that cancels itself out.

3) Even the ones which are interesting to me sometimes contain way more information than I needed.

An example is the Intuitive Introduction of Bayes's Theorem. I selected that one over the technical explanation because I've studied very little math and no statistics. Apparently that's okay; my first answer to the first sample problem was correct. Nevertheless, I kept reading through several rewordings and a long explanation, got the second problem correct, and then started skimming until the theorem itself was presented, which was of course the general form of the contents of my scratch paper. What I learned from that page was some notation, which is handy, but not vital to understanding the contents of Less Wrong--and if it were, I would only have needed to read a small part of the article to learn it. What I actually wanted, a sense of what people mean by "Bayesian thinking" in prose, is actually in the Technical Explanation, where it would not have occurred to me to look. I was explicitly looking for something intuitive, not technical!

This by no means suggests that the rest of the Intuitive Introduction isn't useful. The point is that it wasn't useful to me, and the general exhortation to "read through all the sequences" does not take this into account. (The wiki FAQ is clearer about how necessary it actually is to read every word of the sequences, but I wouldn't assume that everyone reading the sequences comes through the FAQ.) Someone else who had spent more time than I have reading and thinking about, say, transhumanism, before arriving at Less Wrong, would be an even better example. There are some introductory articles that person would not need to read, and for the reasons noted above, it would not be easy to tell which ones.


A solution to these concerns, then, is one which would introduce a new reader to Less Wrong with an brief and unambiguous amount of introductory text, including appealing and clearly-labeled links to further reading. Newcomers could be directed to this and instructed to read until they understood. Here is one way such an introduction could look:

Create a one-page overview of the subjects Less Wrong addressed, packed--necessarily--with specialized language and references to concepts that are explained more fully in the sequences. Hyperlink those key terms to brief glossary definitions; then, from the glossary, link to the relevant posts or sequences.

From reading the overview, a newcomer would immediately get an idea of what Less Wrong is about. Anything they encounter which they do not understand, they can explore at the level of depth which matches their need and interest. Someone already well-versed in Less Wrong's topics could read the one page and jump in; someone who's never thought about any of this before but wants to participate would end up reading all of the sequences. The vast majority in between would be able to fill in whichever gaps they have.

A couple of specific suggestions have been made in discussions about this. Will_Newsome pointed out that one thing that should definitely be addressed explicitly is how each topic relates to rationality, and therefore to Less Wrong. In IRC, chelz observed one hazard of not reading all the sequences: I don't know what I don't know, and just because the title of a post seems obvious doesn't mean it won't contain an insight I could have used. I think this is a good point, and we agreed that the overview would need to discourage this, for example by summarizing what was really addressed in a sequence rather than just the topic described by the title.

It might seem that with this kind of short introduction in place, a newcomer might not read as much of the sequences as experienced participants might prefer. However, my understanding is that that's already the case! I think that for a given reader, this proposal would either make no difference (you can't read less than nothing) or improve the situation by making the same information more accessible.

A sneaky side effect would be taking away most of the reader's ready stopping places. Instead of being able to finish three articles and wander off barely having started, they'd be in the midst of a page until they were done. I suspect that this, as much as the brevity, would increase the number of people who read through to the end.

I'm happy to put my time where my mouth is. I don't think I'm qualified to write the overview, but if the sequences are indeed well-chosen, the list of them will make a good outline for a veteran to flesh out. The style would ideally be brief and practical; putting it on the wiki would allow several good editors to join forces. What I can do is help identify terms in it which need to be defined, as well as write some of the glossary summaries of sequence posts. For those tasks, my newbiehood is an asset, as is my relative dearth of formal education; I won't make the same assumptions about prior knowledge that a long-time reader of Less Wrong (or someone with a statistics degree) might make.

I'm looking for feedback on this idea as well as alternative proposals; when there are a few, or when some time has passed and there aren't, I'll post a poll to evaluate interest. If the result of that indicates getting a project going, it'll need some volunteers who are more experienced than I am with both Less Wrong and MediaWiki.

Comments

( 2 comments — Leave a comment )
(Anonymous)
Oct. 30th, 2010 02:36 pm (UTC)
Re: The Sequences
For what it's worth, I found reading the sequences very enjoyable (even if sometimes difficult), more so than reading the Intuitive Explanation of Bayes' Theorem.

-- Emile
relsqui
Oct. 30th, 2010 05:57 pm (UTC)
Re: The Sequences
Yeah, it's become more clear to me since I wrote this how much it's a matter of "you have to read this" vs. "I recommend this as a way of making something else make sense." I do enjoy most of the old posts I've read, and now intend to read more, but I'm just unwilling to do a bunch of homework for the sake of a website. So I think the "recommended resources" and "favorite posts" lists are a big improvement.
( 2 comments — Leave a comment )

Profile

misc - penguin me
relsqui
I'm not silly, I'm just seriously impaired
Chiliahedron

Latest Month

August 2011
S M T W T F S
 123456
78910111213
14151617181920
21222324252627
28293031   

Page Summary

Powered by LiveJournal.com
Designed by Lilia Ahner