Wikipedia talk:Wikipedia Signpost/2007-07-30/Citizendium analysis

Hi Ragesoss,

Mike Johnson from Citizendium here. I thought this was an interesting article with generally true and interesting facts. I liked the amount of article research and writing style analysis that went into it. However, the general tone of it seems vaguely hostile towards Citizendium (in certain places unfairly so)-- I realize this isn't an encyclopedia article, and perhaps it's none of my business, but perhaps more could be done to make it NPOV?

The main reasons why I feel this article is sometimes vaguely hostile is word choice, the choices made of what aspects and issues to mention and not mention, and the choice of comparison to Wikipedia-6.5-years-in, not Wikipedia-8-months-in. I do, however, believe the article is written in complete good faith(!).

One thing I would direct you to if you're doing research on what's going on at Citizendium is our main mailing list. I think a lot of interesting project discussion goes on and frankly, some exciting plans have been expressed on Citizendium-L. Some of the cool things that'll hit in the next few weeks include

- A semi-automated registration system which should eliminate our application backlog and allow us to add users much more quickly (this should significantly change our numbers-- registration has been our bottleneck) ;

- A lot of really neat initiatives we've started to call "Citizendium 2.0" (hey, not my choice of name. :) ;

- The "Eduzendium" project.

Those are my thoughts. In the end, I see Wikipedia and Citizendium as complementary sister projects (and I really like both of them, for different reasons). The rising tide of getting more people involved in more wiki models lifts all boats. I hope many Wikipedians take that view, as well. --Johnsonmx 15:05, 20 July 2007 (UTC)

I agree, this has a very strong bias against CZ. Without further changes, this should not be included in the next signpost. 71.252.197.151 15:30, 20 July 2007 (UTC)

Mike, thanks very much for your comments. I thought about looking deeper into the mailing list, but I had trouble identifying what was important and what wasn't; your pointers will come in handy, if I have a chance to do more research before I go off traveling (beginning tomorrow). In terms of word choice, which ones are you referring to? I admit it's a bit snarky when it comes the unattributed Wikipedia content, but that's something CZ should have dealt with long ago. Even with the Approved articles, there are more unattributed Wikipedia forks than attributed ones; this points to a systematic problem, an endemic hostility toward Wikipedia. The quote from the d'Herelle discussion page (by David Tribe) is just the most egregious example. The whole licensing discussion seemed to revolve around how much it would be feasible to prevent Wikipedia of benefiting from Citizendium. Your more positive attitude to WP seems to be shared by only about half of your fellow Citizens.

As for the fairness or unfairness of comparing CZ and WP at these different stages in development...I tried to do this as fairly as possible. I did not compare overall site stats to WP at all, but to Conservapedia (which started around the same time). In terms of individual articles, I think it's a perfectly fair comparison. The whole point of creating CZ forks is that CZ editors expected to be able to quickly improve WP starter material while the original articles stagnated or degraded under the open editing of the anonymous masses. In general, that has not happened; most WP articles have improved considerably more since the forks than the CZ counterparts. I don't think that's because of any particular competitive focus on the topics CZ chose to work (i.e., "let's makes sure CZ doesn't beat us"). The CZ Approved articles represent articles that have seen the most collaboration and work, and neglecting the vandalism/reversion cycles on Wikipedia, these parallel forked articles have had comparable numbers of people working on them. CZ can never expect to have as many writers as Wikipedia (unless or until CZ takes over a substantial portion of Wikipedia's mindshare among readers), and that's built into CZ's model, to make up for quantity of writers with higher quality of writers. I didn't go into it, but by my estimate the rate of article editing on CZ is down ten fold from just a couple of months ago.

It's hard to say what attrition was caused by earlier sign-up backlogs (though it was probably substantial), but in the recent past there can't have been that much of a backlog if Larry can now promise 24-hour turnaround. Two new Citizens per day is a stark statistic. That's also down about 10-fold or more from several months ago, by my very rough estimate.

I did not mention the obvious issue of breadth of coverage for related topics or Wikipedia's other advantages of scale. Where CZ has seen success is with writing articles from scratch; there's definitely something to the idea that starting anew gives writers more of a sense of investment, and trying to reform bad content can be more trouble than it's worth. WP would probably be well served to encourage more of that kind of thing. (A total re-write is what I did for my first featured article, Johannes Kepler. Many others have discovered the value of this approach as well, it just doesn't get much emphasis in our guidelines.)

Yours in discourse--ragesoss 16:33, 20 July 2007 (UTC)

I'd agree with Ragesoss that the comparisons are generally fair. The Conservapedia comparison could be strengthened by making it explicit why this comparison was chosen (similar start date, comparable level of initial publicity). Where Citizendium is compared directly to Wikipedia, it's in terms of the development of material coming from Wikipedia originally. That is to say, both projects have the same starting point, an article on date X, and it looks at what the two projects have done with that in the same amount of time. It's not about the fact that Wikipedia dwarfs Citizendium and doesn't rest on that kind of comparison. --Michael Snow 19:32, 20 July 2007 (UTC)

Thanks for the input, Ragesoss and Michael. I see the point about comparisons-- you've got to start somewhere-- though re: Conservapedia, I would suggest that not all edits are equal. It appears from a quick perusal that a lot of Conservapedia's edits have historically been vandalism-related and Citizendium compares very favorably in terms of number of high-quality articles (I admit that's not a very scientific comparison). Regarding comparisons between the content of Wikipedia and Citizendium, well, that's a hard comparison to make. I suggest any devoted member of either community may bring their own biases to the table. Which is not to say nothing should be said about content comparisons-- just that it's difficult to make comparisons meaningfully objective if there's any thread of "us" vs "them" in the comparison (not that I want to fan such flames, but- respectfully- I do get that vibe from the article, though perhaps in that assessment I've been influenced by reading the associated blog post). I did particularly enjoy the writing style & organizational structure analyses you did, Ragesoss, as I mentioned in my previous statement. Frankly, I'd love to see more of it, since to me that's where the really interesting stuff is. Respectfully yours, --Johnsonmx 03:44, 21 July 2007 (UTC)

Gentlemen, one quick thought. This article is not my article, and in extending the invitation to write a counterpoint/survey of what's going on at Citizendium, a lot of my criticisms of bias are muted. One final point that I'd like you to look at is the issue of GFDL violation that my colleague Stephen brings up (short version: both CZ (as evidenced in this essay) and WP (example here are 'guilty' of doing so, which might bear mention if talking about the issue). I'll have plenty to say about edit & contributor numbers-- but I'll do that on my time. Best, --Johnsonmx 13:01, 21 July 2007 (UTC)

---

(after edit conflict:) A few things I noticed when I read the article:

• You're quite positive about the CZ "Life" article; in my opinion it's much worse as an encyclopedia articele than our Life. Their article is more like an essay. An encyclopedia article should (IMHO) start with answering the most important questions, in this case: "What is it?" (the (or a) definition of "Life"). The other question that should be answered in this case: what counts as life, and what does not? (For example: bacteria are life, but viruses are not). The CZ article really does a disappointing job of covering the basics of the subject. Similarly with "Dog": an encyclopedia article on dogs should probably always start with something like: "Dogs are mammals, often kept as pets." Again, the most basic property of dogs, that they are mammals, is not at all mentioned in the article. I had not checked our article Dog before writing the previous sentences, but I'm pleased to see that it does do a good job of giving a basic definition in the fitst paragraph. (See also this blog post, which says it better than I can). Of course, your opinion may differ.

• The CZ Felix d'Herelle article was properly marked as Wikipedia derived, until that tag was lost when the article was moved (see the article's history). The removal of the attribution was accidental, and I don't think you should give it as much meaning as you do.

I think you should make it clear that this article is largely your own opinion. I do agree with most of the points you're making. Citizendium does not seem to be failing, but it doesn't really seem to take off either. It gives me the impression of growing linearly, compared to the exponential growth of Wikipedia. Eugène van der Pijll 16:44, 20 July 2007 (UTC)

By the way, have you seen http://en.citizendium.org/wiki/CZ:Statistics? -- Eugène van der Pijll 16:53, 20 July 2007 (UTC)

Eugène, thanks much. I did note that with the d'Herelle article (and the John Franklin article), the lack of attribution seemed to be a clerical error. But I also found the quote striking and wanted to fit it in. Do you think I should just remove most of that paragraph? I agree with you about Dog; this is another example of the informal, sometimes patronizing tone of CZ articles compared to ours. As for Life, I didn't mean to imply that the CZ article is flat-out better, only that it is more fully developed and referenced. (In terms of ideals, I much prefer the kinds article Wikipedia tries to write.)--ragesoss 16:57, 20 July 2007 (UTC)

The quote on the talk page is interesting, but you cannot judge a whole project by one remark of a single contributor. Imagine what your conclusion about Wikipedia would be... So I agree that your current short mention of the problem is enough. On the other hand, it illustrates the problems of proper sourcing, which can have significant consequences on CZ, as the license of an article may depend on that "wikipedia" tag. (If CZ decides to adopt the GFDL for all their articles, I wouldn't think the consequences to be nearly as bad.) -- Eugène van der Pijll 17:43, 20 July 2007 (UTC)

---

First a note to other readers: the post at the top referred to an earlier draft of this essay, & Sage toned down the language a little in later drafts.

Eugene's link to Citizendium's statistics page is worth following. One datum it reports -- & if I were Larry Sanger I'd be concerned about -- is a steady drop-off in new contributors. Even after ignoring the February spike in new accounts, since April there has been a steady decrease in new accounts, active users, and edits. This is a problem I honestly do not want to see Citizendium have: Wikipedia has its own problem with Expert retention, & I would rather not think that we are doing the best job possible.

A last note: that observation that Nancy Sculerati was deleted is very troubling for the reasons you state, Sage. Have you verified with Sanger that it was deleted & not just moved to an "inactive user" area? (The reason it could not be found might be due to an error in updating the necessary links.) Because it is a troubling point, I encourage you to verify this fact. -- llywrch 17:35, 20 July 2007 (UTC)

I asked Mike Johnson (on his new userpage) about Sculerati. I searched for a moved page or an explanation, but didn't find one. Unfortunately, page logs are not accessible on Citizendium, so I could not tell if/when it was deleted and by whom.--ragesoss 17:51, 20 July 2007 (UTC)

Johnson tells me that Sculerati's userpage was deleted at her request, and that it is not standard practice. It's still a bit troubling that they do not retain a bio, since they do retain her contributions and they rely on the identities of editors. However, it seems more a personal issue and courtesy deletion than a widespread problem, so I've removed that section from the article.--ragesoss 19:56, 20 July 2007 (UTC)

Hit job

This frankly reads much more like a hit job than objective, verifiable fact-based reporting.

Also, if the following is true:

• "seven [approved articles] are derived from Wikipedia articles but unattributed as such"

• after the unmodified Wikipedia articles were purged, many users added up-to-date Wikipedia content without acknowledging the source

I'd like to see the hard facts of this so it can be corrected rather than the mere charge.

Stephen Ewen 21:40, 20 July 2007 (UTC)

All seven articles are identified in the article. You're welcome to compare the articles yourself; in each case it is quite clear that the current CZ versions were influenced to varying degress by the corresponding Wikipedia articles. Several of them have the issue brought up on the discussion pages, and in two or three of the cases (Franklin, d'Herelle, and possibly Metabolism) it looks like simply no one ended up marking the WP bubble (at least for the approved version). It also seems relatively clear that there were many Wikipedia cut-and-pastes after the Big Delete, and that it took some time to mark most of them appropriately (and I have no idea how many were missed aside from the Approved articles).--ragesoss 21:58, 20 July 2007 (UTC)

More

Moreoever,

• your statement about CZ's article on John Franklin is simply in error, see its talk page.

• I fixed CZ's attribution issue on Félix d'Hérelle, and it was a mere clerical oversight; as you can see I explicitly stated on its talk page that it needed to be checked prior placement of the approval template. Could you not have simply done the courtesy of bringing the matter up on the Félix d'Hérelle talk page rather than rush to put it in a signpost article?

Stephen Ewen 22:01, 20 July 2007 (UTC)

This article has not yet been published, and I would have contacted someone at Citizendium before publication had Mike Johnson not found it first. As for the John Franklin article... I overlooked Russell Potter's reply; I find some of Potter's contributions (as User:Profrap) in the WP article history, but it looks like the article evolved organically from the WP version, and Profrap's contributions don't seem to account for the remaining similarities.--ragesoss 22:17, 20 July 2007 (UTC)

For example, prose that originated from this contribution by another Wikipedian is retained in the Citizendium article.--ragesoss 22:28, 20 July 2007 (UTC)

I will look at that, Sage. But again, don't you think the professional thing to do is communicate such concerns on the article talk page or to constables@citizendium.org, which is published clearly? Stephen Ewen 01:08, 21 July 2007 (UTC)

My main concern is to report accurately on the state of Citizendium and the ways it relates to Wikipedia. If some of the problems I've uncovered get fixed, that's great, but I don't have much interest in policing GFDL compliance. This seems to be an endemic problem on Citizendium, and I suspect it goes beyond just the Approved articles I examined. As a quick test, I pressed "random article" until I found an unattributed Wikipedia article. It took ten tries ([1]); the first version was marked WP, but none of the subsequent versions. Another issue is that, even if a CZ article ends up with all the original WP precise text removed, the original structure will often have a strong effect on the result, and should arguably retain GFDL status. Clearly, Citizendium cannot rely on Wikipedians to ensure compliance, as these problems have gone unreported so far.--ragesoss 01:36, 21 July 2007 (UTC)

As for bringing the matter up on the talk page: we can't. Only CZ members can edit there. Eugène van der Pijll 22:33, 20 July 2007 (UTC)

I have a CZ account. (I signed up during the 3 weeks of free-for-all signups shortly before the March opening.)--ragesoss 22:35, 20 July 2007 (UTC)

OK, I haven't said anything then. Eugène van der Pijll 22:41, 20 July 2007 (UTC)

Please see my post at Sage's blog at http://ragesossscholar.blogspot.com/2007/07/citizendium-struggles-to-reach-critical.html Stephen Ewen 05:40, 21 July 2007 (UTC)

Nancy Sculerati

For more info, see this post. That was a while ago though, not sure if it's related to the current event.

Also, the automatic account request software should be on testbed, and then live soon. It is basically an extension of mine I whipped up recently. Voice-of-All 07:34, 21 July 2007 (UTC)

• That post is unlikely to be the reason for her departure; it's from a few months earlier, and the problems she had then (apparently, vandalism by accounts with names similar to hers) are probably no longer a problem. Citizendium has not had any large vandalism problem since the end of the open sign-up trial in February. And it is really none of our business why she left. Eugène van der Pijll 10:03, 21 July 2007 (UTC)