Archive for the 'Online forum structure' Category

An idea for a new forum

I just had this idea for a new way to organize discussions. You’d start out normally, with a short essay stating a position with supporting arguments. Then people add comments disagreeing with certain parts of the essay. Disagreements fall into one of two categories: people who disagree with logical steps made in the argument, or people who disagree with the premises of the argument. (Agreeing/expanding, free associating, joking, or topic changes could all be handled like normal comments.)

People’s disagreements would be responded to by expanding upon the original text, either by editing it, or by adding an explanation that’s linked from the original text. (You could also respond in comments, but there are various reasons why that’s less than ideal. If you’re going to use a system like this, you’ll end up learning when to edit and when to comment.) The link would be in a superscript, like Wikipedia’s links to footnotes and references. The linked explanation would be shown in a column to the right of the original, using a bit of fancy javascript or somethin’. Responses to the elucidation could be shown below it in the new column. (This would end up involving some horizontal scrolling, unfortunately. There are some things that could be done to make this less annoying.) Links to elaborations that have responses could be colored differently so that people that want to just read through the various responses could find them. Also, there could be a link at the bottom of each response to the next response, either in document order, or in chronological order.

The disagreements themselves could be moved from the bottom comment area, to a new section linked to from the part of the text they’re disagreeing with. (This would be more or less useful depending on the length of the text.)

The idea of discussion participants trying to modify position statements in response to each other’s changes, instead of commenting back and forth without updating the original statements, is an idea I learned from Robin Hanson, which he called “reflective equilibrium”. I think it has the potential to make discussions more productive and less error-prone, but making it practical for general use is problematic. In general, the implementation would need to be friendly to ad hoc use by mildly interested participants. If it’s complicated, it’s going to scare people off, and people who do participate won’t be as involved as they could have been with a simpler system. If it has a steep or high learning curve, it’s going to shrink the pool of people who will end up being drawn into discussions. Naive implementations just won’t work.

Truth Mapping) is one such implementation. (I’ll set aside the primary weakness of the site, which is that the initial argument must be stated in terms of the logical steps of the argument. That’s not practical for widespread use, and may not even be practical for more specialized use.) It work like this: you see an argument you disagree with, and you add a comment (critique, in its terminology) stating the disagreement. The author of the argument responds with a comment. At that point, you’re out of comments. All further discussion has to take place by editing those two comments. Now, in theory, this ought to work. Any objections made to your reply could be addressed by changing the original reply to address them preemptively (or apparently so, in retrospect). But in practice, this doesn’t happen. At the Truth Mapping site, most of the arguments I saw (which is quite few—the place has been mostly empty for years) took place by appending replies to the bottom of comments, which defeats the purpose of the format pretty soundly.

Why do the participants resist the RE (reflective equilibrium) format? What makes it inconvenient and unnatural? Is is just a lack of familiarity with the format, or is it something about the way people want to have discussions? If so, can these habits be subverted?

My hunch with my latest idea is that there a certain kinds of inquiries that happen in discussions that are awkward and inefficient with pure RE, that are better served by the back-and-forth of normal conversation. One such inquiry happens as follows: I respond to an argument, the arguer responds, I edit my response to address the points made. But the arguer still feels his objections are valid as they stand, despite my attempts to address them. What happens next? With pure RE, the answer isn’t clear. The only obvious thing to do is to ask the person to clarify or expand upon their response to my reply.

If limited in certain ways, this back and forth might not defeat the purpose of using RE in the first place. But there is a danger that it will. I guess the only thing to do is to implement this system and try it out. (I implemented my last system, but it never got used. Let’s hope this one turn out a little better.)


Unfogged Greasemonkey script

I made a little script for Unfogged comments threads that linkifies numbers in comments to go to the comment with that number. Link.

It sort of breaks on comments with links in them that have numbers in the URL. Not sure how I’d go about fixing that.

It should be extendable, with trivial modifications, to most other blogs with comments that use comment numbers in the same way.


Wordpress category hider plugin

OK, it was so easy to write a plugin as opposed to doing other things that I went ahead and made a plugin to do it.

My category hider plugin allows you to make posts in a selected category (or more than one) display a stub text on the main page of your blog and in your rss feeds. The post still displays in full in your archive pages (but only for the selected category) and on its own page. Of course, tweaking where it gets hidden is trivial.

If you want the category’s posts to be completely absent on the page or in the feed, I doubt a plugin would allow you to do this. If it is possible, it would be a very different plugin. (If you know PHP, on the other hand, the hack is pretty straightforward.)

I’m not going to bother providing a download right now. If you’re interested, comment, and I’ll see about getting together a download for it.


On diarizing

I used to keep a diary. I think it was for about three or four years. Over a thousand entries in the thing. I stopped that when I stopped needing it so much. I stopped it when I realized that no one would ever really like to read it. I mean, even I didn’t like to read it. Just to write it.

But now that I’m not keeping it anymore, now that I have a real, public-oriented blog, I sometimes miss being able to go on and on about personal things without worrying about boring people. So, I wonder if there’s a way to include that sort of thing on my blog without making people who aren’t really interested read them too. I could put the post under a fold (I wish there were a better term for that—I was trying to google it a while back and had a hell of a time finding what I was looking for), so that people had to click on a link from the main page. With a warning above the fold. And that’s all great, except that the whole post would still show up in my RSS feeds. I don’t want to put a password on it—I want it to be publicly available, but only protected from wandering eyes.

I could start another blog, but that’s excessive, and it would dilute my readership anyway. What would be ideal, I guess, would be to have a separate section of the blog. A category that doesn’t show up in the main feed or the front page. Or maybe an author that gets a separate directory. Maybe another wordpress install? I’d rather avoid that if I can, but it’s one possibility. Maybe I should look up a “main page & feed” hack that lets me make certain posts less visible.

UPDATE: OK, so it was easier to write my own plugin. Here.


Why Amygdala doesn’t have commenters

Addressed to the owner of Amygdala. In response to this comments thread.

If you want me to speculate about why you don’t have more commenters, Gary, I can. (And if you don’t, well, you can stop reading now.) You write very a lot of very long pieces. Many people don’t like to read blogs that have over a certain number of average words a day, because it takes too much time to keep up with it all. (Many of these people don’t read very fast.) You could consider splitting your blog into different blogs for different purposes. Political posts on one blog, sci fi and comics posts on another. Something like that. That way people won’t have to make the effort to skip posts they’re not interested in, (and believe me when I say that this has a huge effect—most people don’t have good filtering skills, and will stop reading a blog regularly sooner than skip uninteresting posts) because you’ll already have done that. You could consider cutting down on the word count. Editing posts down more, making them more pithy.

In many posts, you’re basically quoting other people. Quotes necessarily involve a switch in voice and style and context, and so reading them is much harder than reading a single voice throughout a post. The more quotes you include, the harder your posts are to read.

Your posts tend to be “expert” type posts. They show that you have a really large amount of knowledge about what you write about. They often require a certain amount of knowledge about those topics to understand and enjoy. People who don’t feel as knowledgable as they think you see yourself will almost never comment on this type of post. (If they think they know as much as you, but think that you think you know more, they won’t comment.) People who don’t want to make the mental effort to parse large numbers of details in order to understand your posts well enough to comment on them (even if they would be able to, and would have something to say) won’t do so. (This is the reason my blog doesn’t get many comments.)

Your writing isn’t very entertaining. A lot of it is just cursory comments on stuff that you’ve read. A lot of the value you provide is in linking to the stuff in the first place, and while that’s quite valuable, it’s not something that draws a big following, or a lot of comments. (In fact, I imagine more blog owners read you than do people who only comment, since you’re a good content filter to use for material for others.) But most of what people like in blogs is commentary—opinion, argumentation, conversation. Your writing lacks color, stylistic variety, metaphor, and narrative flow that engages the reader. (Yes, that is essential to good non-fiction.)

To the extent that people are drawn to blogs because of the owner’s persona, yours stinks. When you refer to your personal life, it’s always in a negative or apologetic way. This repels people. When you provide people insights into your life, they should always be calculated to endear people to you.

Oh, and you require registration for comments.

That probably covers most of it.


The importance of gradual improvement

A lot of the time I feel like I should just reinvent everything, and fix all the problems with online forums at once. But that’s stupid and silly. The social evolutionary process that has led to our current generation of intellectual interaction software is much, much smarter than I am when it comes to figuring out what actually works. Now, I believe that natural selection hasn’t had a chance to operate on some forum features that would really make big improvements to the quality of discourse. And it may be that there really is no social demand for that kind of quality, and that many of my ideas that would otherwise work won’t end up working. But in order to take advantage of the critical mass of commenters necessary to really test out a new feature, I really have to be able to integrate it with accepted forms of content production. Things that actually do draw commenters, and bloggers. Otherwise it’s way too easy to miss the mark.

Another perk of incremental changes is that it becomes plausible to market the features to existing bloggers with existing communities, instead of having to displace all existing forms of blogging software.

And if I do this, feature by feature, eventually it’ll end up being very close to the ideal. That is, unless there are some strange barriers in the fitness landscape (not very likely, IMHO), i.e. some large upfront costs/risks to some features that would otherwise be sure things. Transition costs.


Subthreads

Reading this thread on Alas, a Blog, gave me an idea. A comment was left disputing another comment that was tangential to the main topic of the post. This happens all the time, but this particular commenter apologized for the tangent (as doesn’t happen quite as often). And, I thought, this happens all the time. Why should people have to apologize for it? What’s wrong with blogs that the make it so tangential comments aren’t appropriate? That really sucks. But there’s a solution.

Subthreads. Have the topic of the main post be handled normally, with comments following serially afterwards. On the comment submission form have a checkbox that says “make this comment a subthread”, and let the commenter enter a title for the subthread. Also have an admin tool to make a particular comment, and maybe some replies, a subthread. A subthread has its own reply-to link, perhaps its own comment submission form, and is colored differently and indented. Perhaps it’s hidden with javascript by default. It appears in the place where its first post would be in a normal comment thread, and all replies are grouped together, followed by the rest of the normal comments.

This would be halfway between hierarchical threading and serial threading, with room for user and moderator participation. I think it would avoid the worst features of both different threading models for comments. Yay me! Now I just need to implement a wordpress plugin for it.


A way to manage comments threads

Being a meta blog as Metablog is, I want to start getting really meta. I know I was going to review truth mapping, but I feel really conflicted about that, and so I’m going to put it off.

In the meanwhile, I’ll introduce an idea really similar to truth mapping that I think could be a good way to handle the problem of comments threads just getting too damn long to read. The author of the post, as part of their authorly duties, integrates all the substantive comments into the main body of their post, as part of a sort of socratic dialogue. They don’t update the post itself as comments are coming in, but keep a copy in draft. Then, when comments have slowed down, they can post an updated version of the post with integrated comments. This can probably reduce the volume of words by 60% and still keep all the substance. It also has an advantage of making things much more clear to everyone, and of not violating commenters’ expectations enough to really discourage participation.

I imagine the integration would proceed by first directly modifying the post to incorporate criticsms and suggestions that the author agrees with. After that, there come criticisms that the author can empathize with, those being stated in their strongest forms as understood by the author. These might be a bit more distracting, so they could be formatting specially, using a special tag interpreted by the blogging software that would both indent, bullet, and hide the text inside the tag. Then to read it, you’d have to click on it to expand the text. Further, each of these little bullets could, optionally, have some explicit logical relationship to others, making it a “truth mapping” sort of approach.

Finally, this improved concoction could be posted “over” the old one, requiring special support from the blogging package, of course. You wouldn’t see the old one any longer, unless you clicked on a revision history link on the post. Most of the old comments would disappear as well. Finally, when the comments veer off in strange directions, they could be split off into new posts. If the author doesn’t really take an interest to a particular tangent, they can simply not moderate that new thread as much. Another blogger could even “take ownership” of the discussion, and comments could be transferred to the other blog where moderation and integration could proceed in the new venue.

There are some comments, however, that don’t really fit in to be integrated into the post, but the author might feel are worth keeping in the latest revision. So it might be necessary to mark each comment as “keep” or “leave”. Or perhaps it would be easier to simply mark certain comments for deletion when their contents are integrated into the main body of the post, and let other comments be assumed to be worth keeping.

All this sounds really labor intensive, doesn’t it? Well, it really is labor intensive. But I think there would be a lot of benefits. Posts would have a lot of reuse value. Instead of having to argue this or that point ad infinitum in some comment thread, one could instead simply link to the relevant parts of the post, or perhaps copy and paste. The results of that new discussion could contribute back to the old post. Futhermore, the bar to contributing to a really old post with a lot of comments would be a lot lower, because one wouldn’t have to read all the comments from old revisions, and so those old threads would stay alive and healthy longer. Plus, it gives more subtle and/or learned commenters more rewards for their efforts, and thus raises the quality of the blog’s commentariat.

This approach would require quite a bit of programming beyond what I’ve seen available in any system (except maybe Drupal), but I think its potential is large.

More than more explicit and formal approaches to debate modeling, I think this approach has the potential to catch on in the wider blogosphere, at least at the more serious sites. And it will give us the kind of experience and intuition and ideas we need to really know what we would want out of a more explicit system. So it’s a good first step.


Truth Mapping

Wow, even I wouldn’t have picked a name that corny. And what a terrible color scheme.

But it looks like they’ve come from exactly the same perspective I have. Hat tip to Will Pearson from the SL4 mailing list. I’ll comment later.


Working on Debate Modeling

In lieu of a bunch of mindless rambling, I’m going to spend {whatever time today I spend on Debate Modeling, and not watching the Chronicles of Narnia}*, dissecting this this typical blog discussion to try to figure out what sort of elements can be atomic enough and yet still capture all of the good content (as opposed to the unwanted content). Actually, it’s not an entirely typical discussion. It has less noise, and more controversy. It does have a few people playing tug-of-war about presenting evidence, each presuming dishonesty on the other’s part, but there’s only one thread of that, and it’s pretty explicit, and I need to think about that stuff anyway.

* These braces are generic phrase grouping markers. Not standard english, but oh so useful in writing long, nested sentences. They’re not needed often, but when they are, it’s really annoying to rearrange things to work with The Man’s English instead of just using them. You don’t want to support The Man, do you?