Article drive strategy!

[Larry Sanger]

Martin, Lee & Aleta--thanks   

I do wish someone would engage the specific argument I made, the one that explains my resistance to you credit hogs.     You know--I won't say it again, but I mean the one about how this runs the risk of reducing the degree to which the project is collaborative, and thus damaging the essential thing that makes the project go in the first place.

[Martin Baldwin-Edwards]

I agree, Larry, but there are possible solutions to that. We need to have a technical means to determine the main contributors to an article: sometimes it is absolutely clear, from the history, that it is one or two people. On other occasions it may be more complex, and in some cases it may be so multi-authored that it is impossible to say anything.

What we need is a rule, so that if [for example] 3 people have contributed more than 25% each, then they should be named as authors. What I am thinking here, is that anyone can still participate in writing, but there would be an incentive to actually participate in a major way for specific articles -- that is, to develop them and take credit for the hard work. On the other hand, if someone likes to add small things to a lot of articles, s/he is still able to do this, but will not nornally be credited.

How does this sound, in principle?

[Aleta Curry]

...
I do wish someone would engage the specific argument I made, the one that explains my resistance to you credit hogs.     You know--I won't say it again, but I mean the one about how this runs the risk of reducing the degree to which the project is collaborative, and thus damaging the essential thing that makes the project go in the first place.

Sorry, Larry.  I thought this was done.  If you'll point me to the thread where you made your comments, I'll respond specifically...point by point...whatever it takes, because I think this is important.

I'm not at all sure why one thing has to lead to the other.  If the collaborative nature of the project is in any real danger (as opposed to the speculative Nth degree extreme Americans tend to argue in) then why on earth can't we take specific measures to limit the possible damage?  Indeed, I think these concerns are mitigated by the fact that we've got some problems with the project for which we simply must find solutions if we are to be successful.

[Larry Sanger]

...
I do wish someone would engage the specific argument I made, the one that explains my resistance to you credit hogs.     You know--I won't say it again, but I mean the one about how this runs the risk of reducing the degree to which the project is collaborative, and thus damaging the essential thing that makes the project go in the first place.

Sorry, Larry.  I thought this was done.  If you'll point me to the thread where you made your comments, I'll respond specifically...point by point...whatever it takes, because I think this is important.

I'm not at all sure why one thing has to lead to the other.  If the collaborative nature of the project is in any real danger (as opposed to the speculative Nth degree extreme Americans tend to argue in) then why on earth can't we take specific measures to limit the possible damage?  Indeed, I think these concerns are mitigated by the fact that we've got some problems with the project for which we simply must find solutions if we are to be successful.

I've made the point several times I think, but this is the most recent.

[Lee R. Berger]

Ok - One by one (by the way Larry - don't feel frustrated - I also think I have answered these before one by one but now we are all in one place!)

Sanger quote - "The one thing that keeps this project going, the fuel that makes any genuine strong collaboration run, is the sense that we are all equally empowered to work on any article.  To put my cards on the table, I have had some experience--with the Encyclopedia of Earth--designing a system in which contributors are credited.  I wish I could give you a report based on more recent experience (I'm no longer associated with that project), but as far as I can ascertain, the result has been that project never really generated the sort of mutual excitement that comes with strong collaborations.  We are typically happy and impressed when others edit our articles, and the sense that we're working on something much greater than any one of us could create makes a huge difference to the degree of motivation people feel to write.  My worry is that, if we start crediting all the collaborators of an article, and especially if we are selective about which collaborators we credit, we will set up a system where others, who aren't on the list, will feel they have to get permission from those who are on the list, if they are to do anything with an article.  I am fairly sure, based on my experience with different wiki systems, that this is the case."

Lee's reply - Ok - its hard to argue with your experience, but since we are trying to create something that is different than what has gone before, those models may not be applicable.  They may be useful to examine, but may not apply exactly.  Having said that, let me attempt to address the two core points of your argument:

1) Larry's comment - The one thing that keeps this project going, the fuel that makes any genuine strong collaboration run, is the sense that we are all equally empowered to work on any article.

Lee's comment - Yes and no - of course we can all edit away - if we have stated our real names etc. but you have drawn different levels of access and rights to different levels of people.  And with the approval system (and we really must go back to that topic sometime soon) and with no clear replacement draft policy - one cannot just "edit away" in an approved article.  I would suspect - although I'm willing to be proven wrong - that once an article is approved and we have our general act together, the editing to that article will fall way off and it might get updated once every six month as a new discovery, new knowledge etc. is created.  CZ is very different than Wikipedia in this respect - stability of an article once it has reached maturity is probably a goal here (thus my CORE article arguments - which we must get back to.....)

and

2) Larry's comment - My worry is that, if we start crediting all the collaborators of an article, and especially if we are selective about which collaborators we credit, we will set up a system where others, who aren't on the list, will feel they have to get permission from those who are on the list, if they are to do anything with an article.

Lee's response - My proposal - as has been stated - was SPECIFICALLY that we credit all contributors - they are there anyway!  I think it was you or others that were suggesting that we only credit the first author etc. All one has to do is look at the history list and see who wrote what.  The very fact that a publically available history list exists (even on wikipedia, which doesn't seem to have the problem you bring up here) argues that most people don't think like this. The difference between just having a history list and what I and others are suggesting is that somewhere AWAY FROM THE WORKING PAGE - a formal acknowledgement of contribution is present.  I suggested in a previous email that this might described as Major authors, approving editors and contributors (those who just dash in and make an editorial change etc.).  It defeats me as to how this would run people away?  My gut reaction is "contribute" - get on the list!  Isn't THAT actually what drives most of the group participation and logarithmic growth that makes these wiki concepts work?  The only exception to this that I can see - and here is where you have a point - is what happens when an expert comes along and really feels the approved article - written by x - is crap, wants to do a big rewrite/addition but is nervous about offending x (maybe they are in the same field, know each other, date each others wives etc.).  In that case we should have some "how to report major errors/corrections anonymously policy" wherebye a person could declare themselves (their real names) to an editor but ask for anonymity (just the way most journals work). The editor could then, if the case merited it, forward suggested changes to the author and monitor the progress of correction (or call in constables, other editors etc. to help make right). end of Lee's comments on this.


Larry's comment - I doubt we'll have this problem if we credit people only with the first draft, and if the credit is given on a special page, separate from the article itself.

Lee comment - I just don't see the logic in this.  As we have discussed, the first draft may be the simplest thing to do - it is also NOT the goal of CZ - we are here to have 1,5 million APPROVED articles.  To get there we have to attract experts to drag stubs and developed articles to that stage.  If they don't come - en masse - to CZ I don't think we'll build the critical mass to do this.  The ONLY way I can see that you could attract busy experts is to offer "credit" - this credit can be hidden away from the public eye - It could be in a "Contributors page" against articles.  Anywhere, just so a promotions committee, supervisory, etc. could go to the article, find a link and check whether Joe smith did what he said he did. Yes, I am biased towards academics in this argument, but surely they make up the perceived bulk of potential contributors to CZ?  They write for a living, are experts on specific topics, are trained to edit and often have the time (I didn't just say that out loud) AND they have a vested interest in sharing knowledge IF it contributes towards promotions, publications and prestige. I know some of you don't give a sh**t and some will think this is an awful shallow assessment of modern academia but I believe it is a reality that cannot be ignored. Sharing information - for credit - is what we do. If not, why then did Einstein not just submit Relativity anonymously? End lee's reply

Larry's comment - If we were to credit all Citizens who have made the slightest little change to an article, I think that might be all right.  But I'd still be, quite frankly, worried about the slippery slope problem.  Already I think that our contributors (just as on Wikipedia, to be fair) sometimes feel a bit too much personal ownership over articles they are creating--they sometimes "shoo away" people who arrive to help.  This doesn't happen a lot, but it does happen.  Authors who see other people, whose contribution to an article wasn't nearly as significant as their own, getting "equal credit" in an alphabetical list (or a list by order of first edit), will in time feel that they need more distinct credit, if names are named at all.  I can even see people starting to resent the contribution of newer people, as it dilutes their own personal credit.

Lee's reply - May we get to a stage where we have enough editors and authors of APPROVED articles that they feel possessive about them and shoo people away.  As I argued above, the additions to an approved article are probably going to be minor anyway if we do our job right - if they are not minor then put a system in place for anonymous address of the issue.  For more incomplete articles this is a problem but probably nothing that couldn't be handled in the normal course of events UNLESS they were CORE articles - then this sort of behaviour would require intervention - but if we get those out of the way and approved then that's not much of an issue either. Also, many of us are suggesting credit by SIGNIFICANCE  of contribution - certainly not in order of contribution or in alphabetical order - that would make no sense!


End of Lee's commens (thank goodness you say:-)

[Aleta Curry]

I agree with Lee's treatment of this, and agree with his conclusions but my initial reaction took a different tack:

Larry, I certainly understand your concerns, but I don't find your arguments convincing.  Specifically, I do not think that the examples you use prove your point, indeed, they present evidence to the contrary.

If I understand you correctly, your fear is that credited articles lead to a breakdown in the collaborative process, yes?  I didn't see that when you first brought it up, and I still don't.

Item 1.  A proprietary feeling about an article--a sense of ownership of the work--this is mine, keep off!  Shoo, Shoo!--may occur and is to be avoided.

Yeah it happens.  Hands up anyone who's never felt this way in the slightest-even about a measly paragraph of text?  Anyone?  Tell the truth!  Nah, I didn't think so.

You quite rightly point out that this is undesirable in our project, BUT you also point out, again, rightly, that this happens both here and at Wikipedia.  But, Larry, NEITHER of these projects have signed articles, so the most you can say is, yeah, well, see, see--if even these projects have ownership problems to any degree, those problems will only get worse if we have signed articles.  I'm sorry, but that's conjecture. 

The problem of overly-jealous guarding of articles can be successfully combatted.  First of all, we have a community culture that deplores it.  Lee has made some good and concrete suggestions, as have others.  We can argue about the details of the mechanics later--but why throw the baby out with the bathwater.

Item 2.  Wikis that have signed articles grow slowly or stagnate.

The examples you use here are Scholarpedia (probably wrong spelling) and Encyclopedia of Earth.

I'm hoping I don't describe these incorrectly, but even if I mix up elements of the two, my arguments remain the same.

Scholarpedia has very distinguished scholars who sign their articles.  There's no such thing as a non-academic expert; you must not only be a scholar in a pure academic field, you must be known in the field; distinguished among your peers.  Ch* A'mighty--that's enough of a reason why it grows so slowly and is not particularly collaborative--first, the pool is a fraction of what it should be, second, you've got the "names" of the academic world writing their own articles about their own subjects.  I also think I have read where Dr L Sanger says that Scholarpedia does not operate as a wiki, it is rather expert essays hosted at a wiki, which is quite a different thing.

So the fact that the articles are *signed* has nothing to do with the sense of ownership.  I submit the problem is NOT that the authors get credit for their articles (and by the by, that's probably the big incentive for drawing the big names), the problem is that they are THEIR ARTICLES to start with!

I would make a similar argument for EoE.  It's got a small and specialised focus, at least at present.  I think I signed up with DU ages ago, but didn't hear back timely and it all seems to be about Earth Science, so I didn't see a place for myself.  An outsider cannot explore the mechanics of writing at EoE, so we can't check the article history and see how collaborative it is, but just given the little I can see I would suggest that again, the problem is with the nature of the community, NOT with the signatures.

Larry, I see absolutely no reason why we can't come up with a better model.  None.  There are enough brains around here to make a good plan.  And, if it doesn't work, it gets jettisoned.

The fact is, and it pains me to say so, things are not working in our community as they should be.  Why?  Many reasons, and they're beyond the scope of this discussion.  But for this discussion, motivation and reward seem of paramount importance.  We've already discussed the whys and wherefores of these, so I will only say that pretending they aren't important is a recipe for failure.

[Stephen Ewen]

It's very simple, Larry.  You just have to look at it from the underside.  A Wikipedian, "Mojo", posted on the Citizendium blog something that I am arguing is very widely held sentiment:

It’s nice that Citizendium has experts for quality assurance, but I don’t think it’s a major selling point for rank-and-file contributors.... The real names requirement is a reasonable safeguard, but I doubt that ‘elders patrolling the bazaar’ ranks high on their wish-lists. (See http://blog.citizendium.org/2007/09/09/confused-about-citizendium/#comment-35040 )

We need to give incentive to authors--lots and lots and lots of them.  The current CZ system:

1) requires real names, which is a risk to many; and,

2) after already placing the hurdle of item 1, we insert authors into a system where (to condense it all way down) they must concede to editors in the very task authors are asked to do: write articles. 

Simply, item 2 is not enough incentive--and to a great many it is a disincentive, in fact--to overcome item 1.

The solutions are:

1) relax on the real names policy for authors; or,

2) build additional incentive into the system that will at the same time make both item 1 and item 2 much more attractive. 

That incentive?  It is

3): Main contributors to articles approved by editors get credited by name within the article. 

So it's like this:

You must use your real name.  That remains a barrier.  However, what we trade in exchange is what will give people the incentive to leap over that hurdle: you will, by name, receive credit for your work authoring general encyclopedia articles that have been approved as reliable and accurate by real-life credentialed experts.  That makes working with experts take on real meaning to a broad number of authors, who are having to hand out their real name.

Bottom line: Item 1 and 2 is not enough.  Add item 3 and it may well be.

[Kim van der Linde]

It's very simple, Larry.  You just have to look at it from the underside.  A Wikipedian, "Mojo", posted on the Citizendium blog something that I am arguing is very widely held sentiment:

It’s nice that Citizendium has experts for quality assurance, but I don’t think it’s a major selling point for rank-and-file contributors.... The real names requirement is a reasonable safeguard, but I doubt that ‘elders patrolling the bazaar’ ranks high on their wish-lists. (See http://blog.citizendium.org/2007/09/09/confused-about-citizendium/#comment-35040 )

We need to give incentive to authors--lots and lots and lots of them.  The current CZ system:

1) requires real names, which is a risk to many; and,

2) after already placing the hurdle of item 1, we insert authors into a system where (to condense it all way down) they must concede to editors in the very task authors are asked to do: write articles.

I think you are right on this.

So, can we use the reality of WP to unravel what is gong on at CZ? Yes. The equivalent to approved articles here is roughly featured and good articles at WP.  WP has 2 million plus articles, and haw many good ones are in between? A minute fraction of the total.

Why? Because people are not interested in it. People are editing to get exposure. Not tomorrow, NOW! They want to get the feeling they contribute something, that they know something, they want recognition.  And that is that what you write is read by others. That is not that what you write is maybe approvable in due time when you have put lots and lots of effort in a single articles to get it going in all details in sufficient length dealing with each and every nasty subsection you have no clue about.

So, what you see now is that people do exactly what happens at WP. Write a stub, a small article and than leave it to others to continue. Most of those articles will take a long long time before they will get approved, if ever!

So, if we want really more articles, we have to change two things. Question, how to solve this? We cannot relax the requirement for real names to be know at least with the constables, because that is our main safeguard against vandalism. With the approval process, it would require a pretty much complete different mindset to deal with that. You have to go from complete articles and a rather formal approval process to a  much more reactive response team like process in which anything (small article, section, additional sentence) that is well-written can be approved immediately by any editor with the knowledge on that topic (which can be reversed if another disagrees).

[Martin Baldwin-Edwards]

I agree with almost all of the last few posts. IN fact, this is a very important discussion about how CZ could and should work. My own feeling is that we cannot emulate the breadth of WP, for a variety of reasons, but we should aim for

(1) a reasonable coverage of core articles

(2) the quality of such articles should rival EB

(3) some more superficial coverage of less important articles.


If this is correct, then it implies identifying the core articles for CZ and rapidly taking them to completion. This has to be done at the workgroup level [I notice, to my chagrin, that Anthropology are actually doing that!]

[Larry Sanger]

Let's move forward on this, then.  I am going to take an hour out today and get a resolution started.

[Larry Sanger]

No resolution is really necessary.  I've put the initiative before the Editorial Council for approval "by acclamation"--I doubt anyone will have any objection.  I've written up the page

http://en.citizendium.org/wiki/CZ:Core_Articles

and I think we're ready to go.  Dive in, please!

We can credit co-authors later, if we decide that's a good idea...the whole issue of taking credit vs. robust collaboration really needs to be put to a vote and, perhaps, piloted.  I'd appreciate it if anyone would take the initiative there, too...