CZ:Proposals/Pilot to allow Citizens to take credit for pages

This proposal has been assigned to the Editorial Council, and is now in the Editorial Council proposals queue.

Complete explanation

Driver: Larry Sanger

Here finally is a real proposal for what so many people have asked for, engineered, but never actually put in front of the Editorial Council. We would, to a limited extent, allow Citizens to take contributor credit of articles. My aim here is to articulate a proposal that I'm willing to try out (not necessarily to make our permanent policy) and that we can make into an Editorial Council resolution.

The purpose of this sort of proposal, at least as made by me, is not to allow people to claim credit on a CV. It is, instead, simply to give contributors some modest thanks for their work on a specific article. Many of us think that this will serve as a small inducement to write and edit our articles. Also, if we credit the approving editor(s) in the same place, we give them some motivation to people to approve articles.

The contributor list would (for now) take the form of a template placed at the bottom of article pages. Here is the template, with just three names.

{{Authors|Martin Baldwin-Edwards|Meg Ireland|Russell D. Jones}}

Nothing displays!

Here's the template, with five names:

{{Authors|Martin Baldwin-Edwards|Meg Ireland|Russell D. Jones|Hayford Peirce|Warren Schudy}}

Here is how it displays:

Authors [about]: Martin Baldwin-Edwards | Meg Ireland | Russell D. Jones | Hayford Peirce | Warren Schudy CZ is an open collaboration. Please join in to develop this article!

Features of the template:

• The list is labelled "Contributors" and, if the article is approved, will include the names of the approving editor(s). [The template does not yet do this: if anyone would like to add that functionality, please do. Please try to keep the size of the template small, however: you might want to hide unexecuted code in a separate template.]
• Either small and unobtrusive print, or a by-default collapsed "infobox" style table. Perhaps the former will be used until the list gets too long, then the "infobox" kicks in.
• Lists the contributors to an article strictly in alphabetical order.c
• Names will appear only if there are five names in the list.
• Citizens may add their own names to the contributor template. They are instructed to take credit only if they believe they have contributed at least two substantive sentences to the article.
• One Citizen may list another Citizen on the template, but if a Citizen ask on not to be credited, his or her wishes must be respected.
• If a contributor does not want to claim credit, then it should be possible for contributors to add the words and other contributor(s) to the list of contributors, e.g., by simply writing |other [the template does not currently permit this: anybody want to fix it?]. This in effect means that, even if there are many people who do not want to claim credit for a specific article, there must be at least four contributors who do wish to claim credit, for any of them want to receive credit.
• One may not argue about a Citizen's claim of credit. The Constabulary may edit author lists and is responsible for settling any objections to a particular author listing.
• There would be a small notice wherever the template appears that pithily conveys the notion that, despite our having listed these names, the article is wide open and available to work on by any Citizen.
• Contributor templates work precisely the same way for subpages, that is, each subpage uses its own contributor template.

I propose that we do a pilot project for at least one month, maybe two or three, in which the template's use is limited just to the Biology, History, Anthropology, and Music workgroups.

Reasoning

The central argument for allowing Citizens to take any sort of on-page credit (they already have credit for their specific edits in the edit history) is that this would motivate people to do more work. I don't know if this is true; I actually doubt it quite a bit. But I am willing to try it and see.

Why not go "all out" and distinguish amounts of or roles in contribution?

There is one central argument against allowing Citizens to take credit, and that is that Citizens who get credit on pages will be encouraged thereby to regard the article as "their turf." They will be much more likely to defend their particular edits and versions and to receive new collaborators coolly, if not with outright hostility. There is a perfectly good psychological/economic principle at work here: when you start rewarding people for doing something, they begin competing for the reward by doing it more and more; but they can and often will try to increase the value of the reward by "eliminating the competition." Moreover, "new players" in the "game" are dissuaded from entering (i.e., from joining the roster of authors), because the competition makes doing so more difficult and unpleasant; if one person or a few people look "dominant," the new players cannot get a toehold in without much more effort than he would have to expend under the current system.

This is not just theory. It is how the real world works, and it is absolutely predictable that, under certain versions of the authorship credit proposal, erstwhile collaborators will start competing to be the "top author." They would then have an incentive not only to compete to improve the article, but to downplay the contributions of others. You can already see some salivating at the prospect of this happening, among the various discussions that have happened: people want to have the opportunity to be recognized as the "top author." That would completely undermine CZ's collaborative ethos, replacing it with a very nasty competitive one. If you think that talk page disputes are a problem now, you ain't seen nothin' yet.

Certain features of the proposal are, therefore, absolutely crucial, if the proposal is to be consistent with the open, collegial, and collaborative community that we have set up and wish to build. A crucial element is that we make no distinctions between any contributors for purposes of giving credit. As soon as we do so, we are on our way down the slippery slope: if you eliminate tiny copyedits from credit, then why not eliminate one or two sentence edits? (Actually, the proposal above stands at the top of that slippery slope, since it does eliminate tiny copyedits--on grounds that this will dissuade "credit hounds" from changing one comma and then claiming to be a co-contributor.) Or, at least, why not credit people according to amount of work, or the roles they've played in crafting the article? The answer to such questions is obvious to me: because it will be a disincentive to people to collaborating, and sharply increase the level of competition and hostility we experience in collaborating. I believe I've already seen evidence of this on The Encyclopedia of Earth, which I helped get started in 2005, and which does in fact credit people with roles in the articles. As a result, most EoE articles only have one or two authors, and that project does not enjoy the sort of robust collaboration that we enjoy.

Therefore, I would make the following test for the pilot project: if at the end of the pilot project period, there are very many or loud calls for rewarding people in differing amounts according to how much the contribute, that means the pilot project fails the test. That would be an indication that the community is lusting after individual credit at the expense of the very thing--collaboration--that makes a wiki run.

In fact, if we see many such calls during this proposal period (and I see a few below), I will cancel the proposal, as the reaction will have demonstrated to my satisfaction that the community will become competitive--or, perhaps more likely, more apt to stay off each others' "turf," keeping a polite distance, and asking for "permission to edit" even more than they are now doing. I strongly prefer and prize a vigorous collaboration; this, the very engine that runs the wiki, is far more important than giving people credit for their work on the article page, which is at best just a another way to motivate collaborators for whom personal credit makes all the difference.

Why a minimum number of five contributors before credit is given?

Let us consider again the reason we might have for giving authorship credit. This is absolutely crucial to the disputes we are having over the shape of this proposal. The reason is quite simply that we believe that people might (for a whole variety of motivations) be more willing to work on CZ articles if they are specifically named as authors. But, as I have argued ad nauseam, raising the value of authorship in this way might induce some people to jockey for position (and thereby induce many nasty controversies), and would probably induce many more people to politely keep their hands off "other people's articles." Either of these effects would do far more damage to the project than the (I think) relatively small motivational benefit caused by giving people credit. The only reason I am able to support any proposals at all along these lines is that I can imagine policies that would minimize these problematic effects:

• We do not make any meaningful distinctions among authors in terms of order, whether names are bolded, size of print, etc.--hence there is no incentive to jockey for position and to "shoo away" other authors.
• Authorship information is underemphasized with small print, perhaps behind a (by default) collapsed box, sending the message that Citizens are free to join a loose group; and we state this message explicitly in the authorship box itself, too.
• Names appear only if there are five names in the contributor box, thereby explicitly denying the first contributors a sort of "squatter's right" which might lead them to resent and shoo off "newer" contributors; this also actually rewards collaboration rather than competition.

It is worth examining more closely what the minimum number of contributors should be, before authorship credit is displayed. There have been some calls for reducing the number in the proposal to one or two, but if anything I am inclined to keep the number at five. It should help to explain why credit to one single person should be denied: if people are given credit for contributing to other articles, where there are multiple contributors, why should they be denied credit when there is just one contributor? This looks unfair. Well, in many cases, articles have just one contributor for a long time. This means that, if we credited authorship of articles with single authors, a fairly large portion of CZ articles would appear to be written by just one person. This would, I think, discourage others from contributing to those articles, because this looks presumptuous: if I add just a sentence or two to so-and-so's fine article, I'm suddenly a 50% co-author? This is an aspect of unduly distinguishing individuals, and being the first to contribute is going to look like a distinction. You might say, "And it is one, and it should be treated as one! We love people who start new articles for us!" But I disagree, for the reasons stated above. We precisely do not want to make authorship distinctions among people, because this is tantamount to "messing with" the culture of the project.

My other concern is that there are frequently people who work together, and whose names would appear as co-authors on many articles. What I fear is that a cliquishness might come from this. Again, it will look presumptuous if some undistinguished, new, or relatively unrespected author, X, starts editing--and thereby adding his name to--articles that two relatively distinguished people, Y and Z, have been working on. This will make it look as if X had joined Y and Z's clique, something the latter may resent.

I would expect these and similar problems to arise due to an implicit "culture of credit"--i.e., the academic culture of the fair repayment of work done in terms of the honor of differential authorship credit, with all the problems that attach to that practice--rather than a "culture of collaboration." But we could avoid the appearance of having a culture of credit, I think, by crediting people only if there is some minimum number of authors. Generally, the minimum number should be whatever is necessary to avoid implying that we do have a "culture of credit." What number is that? I'm not sure, but for the above reasons, I think it's at least three. Probably it's more like four or five. Then, if someone adds his name to the roster of authors, it will not seem to "matter" so much--it will not significantly detract from the contributions that others make, because their own "recognition" is already small at best. So I am inclined to make the minimum number five, for that reason. I would encourage those who object to this to direct themselves to the argument behind it, rather than making the facile point that "it's unfair." It's not unfair if the rules apply equally to everyone and if the community has more of an interest to keep the project robustly collaborative.

Besides, this would give us a very positive new sort of goal to aim at: increasing the sheer number of people at work on articles.

Why not restrict credit to approved pages?

It has sometimes been suggested that we restrict credit to approved pages, the idea being that this would provide an extra incentive for authors to improve articles and move them toward approval. After all, if we have worries about the impact of crediting authors on the robustness of collaboration, shouldn't we opt that way? It seems that people would be less competitive about authorship if it were recognized only after approval.

There is an interesting tension (if not outright contradiction) in this suggestion, however, as follows. On the one hand, it is suggested that the "carrot" of authorship credit would motivate people to move articles toward approval. On the other hand, it is said that the lack of authorship credit for pre-approval articles would mitigate any negative impact on collaboration, at least in the pre-approval period. But I think that to the extent to which authorship credit does act as a major incentive, that will raise tensions and competitiveness on pre-approved articles. Indeed, if authorship credit is regarded as a relatively rare reward, that (basic economic principle again) raises its value, which raises the probability that people will fight for it.

This is why I am somewhat opposed to restricting credit to approved pages: it makes authorship special and hence something that is worth fighting over. A forteriori I find myself opposed to plans that would have editors portion out credit. You can easily imagine the disputes and ill will that would follow whenever an editor "got it wrong." And you can also imagine that editors would have one more incentive not to do the hard work of approving CZ articles.

What's the point of crediting authors if it isn't a distinction?

In reaction to this discussion, someone might object, "But what is the point of crediting authors at all if you are designing the system so that authorship isn't a distinction? After all, it is precisely because authorship is a distinction that people want credit for it and would be motivated by it."

Having one's name on a list of authors is a distinction, because it is something one claims that others, who have not contributed, cannot claim. In other words, there is still a significant, meaningful distinction between authors and non-authors, even if (by careful design) there are no distinctions among the various authors. Indeed, there are no doubt some "credit hounds" (I do not mean anyone in particular) who will regard even this small distinction as significant enough to motivate them to make many small changes to articles. A credit line is a small price for the community to pay for their laudable, if somewhat selfish, behavior.

"Still," one might maintain, "an alphabetical list of people who have merely done some small amount of work on the article seems to have no point. So far from being something to be proud of, by listing the leading author of an article right alongside someone who added just a few sentences actually seems to disrespect the contribution of the leading author." This is an understandable concern, but I think it is ultimately not as much of a problem as it might seem to some. Please let me explain.

I want to make two relevant observations. First, by displaying only relatively long, alphabetically ordered lists of authors, we send the clear message that collaboration among equals is more important to us than singling people out. Upon reflection, this cannot be taken to indicate disrespect of leading authors, because it only indicates our priorities. Indeed, there is a certain kind of honor in being the most important person in any collective endeavor, but not singled out. This might strike some observers not as disrespectful but as charmingly modest--especially if the policy is applied consistently across all articles.

But, second, isn't it still pointless simply to list collaborators in this way? Well, it's not pointless if we can find a point; and there is a point, namely, we give some small recognition to the people who have cared enough to read and make at least a small contribution to the article. Granted, while this is a small honor, it is not nearly as big an honor as being recognized as one of the "lead authors" of an article. And indeed, as a community, we are opting not to single out and honor people for their (implicit, de facto) roles as lead authors. But it would dramatically change the collaborative nature of the wiki if we were to recognize certain people as "lead authors" of our articles, as I've explained at length and as seems to be the case on the Encyclopedia of Earth as well as Scholarpedia. So, yes, there is a point to listing collaborators in this way, although it is admittedly not as big an honor as being credited with lead authorship; but there are excellent reasons not to specifically credit people with lead authorship.

Besides, we might well discover that people are actually very motivated by even this small extra honor for their work on an article. It is worth testing whether this is the case before we consider any dramatic change to our collaborative culture.

Issues of disagreement

In discussion of this proposal, various people have raised a number of substantive issues of disagreement. For the benefit of decisionmakers, here is a list of those issues. For this Editorial Council proposal, it is possible for the Council to vote on amendments to this proposal and thereby attempt to satisfy these disagreements. The version submitted is the version championed by the proposal driver (Larry Sanger). Here is a summary of sentiments in disagreement:

1. We should not give this sort of contributor credit at all.
2. We should increase the minimum amount of text to 5%, or 250 words, or something like that.
3. Credit should be given only, or only at first, for approved articles.
4. We should give credit in proportion to the amount that a person contributes (e.g., bold the names of the leading contributors, etc.).

Implementation

Once this proposal passes the Editorial Council, it would be implemented by:

• Finishing designing the template, if necessary.
• Setting a specific date to end the pilot project.
• Creating a page that specifically details contributor list policy (as outlined in the Editorial Council proposal).
• Explaining to the Constabulary what their role in the pilot project is.
• Announcing it using various CZ communication media, including cz-biology, cz-history, cz-anthropology, and cz-music, and asking people to start testing out the system by giving themselves credit!
• At the end of the pilot project, a collaborative report is prepared. This should address the various issues raised in earlier discussions.

Discussion

Archived discussion here

Wipe this clean please

It does seem to me that this type of discussion is something we should learn from and strive earnestly to avoid. This is still a young project, shouldn't we be thinking of testing out ideas as pilots and then deciding, with the benefit of a bit of experience? Would it be so hard to test this on one or two workgroups for 3 months say and then discuss whether it should be extended, modified or abandoned? Would it be so very bad to get it wrong? Some of us may well recoil at how it works in practice, others might be reassured or enthused. If I haven't participated here on this topic it's basically because I'm not that good at reading tealeaves. But probably opinions with the experience of a pilot would be better focussed and informed.Gareth Leng 13:44, 12 March 2008 (CDT)

Previous discussion archived. --D. Matt Innis 14:44, 12 March 2008 (CDT)

REINSERTED MY DELETED COMMENT I will not tolerate deletion of my comments. This is becoming a common practice, apparently, on this page.Martin Baldwin-Edwards 18:22, 12 March 2008 (CDT)

It appears I accidentally deleted your comment when making my own. I have no clue why this happened; it certainly was not intentional. Sorry. Warren Schudy 18:41, 12 March 2008 (CDT)

OK. Apologies for my paranoia about this. Martin Baldwin-Edwards 20:04, 12 March 2008 (CDT)

Although I agree very much in principle with the concept of pilot studies, that agreement is largely predicated upon the idea of alternatives being tested. We know very well from human history that initial structures tend to be retained, unless and until they are completely discredited. Even partial failure leads to modification rather than rejection. Thus, the starting point of any policy will to a great extent determine its endpoint: this, I suppose, is the intuitive response of many of the strong positions that were adopted. It is, therefore, imperative that different pilot studies are carried out simultaneously in order for the arguing on this issue not to recur. Martin Baldwin-Edwards 16:57, 12 March 2008 (CDT)

Martin, did it get moved to the archives or just deleted? --Robert W King 18:38, 12 March 2008 (CDT)

What happened to it warning about edit conflicts? I submitted a comment and it deleted Roberts without telling me. Warren Schudy 18:46, 12 March 2008 (CDT)

I bet the problem was I was viewing an out of date page when I hit edit. The feature that clicking "edit" when viewing the history reverts to the version viewed may be useful but is rather unintuitive to me. Warren Schudy 18:52, 12 March 2008 (CDT)

I noticed that Martin's comment disappeared when Warren edited but decided not to be a busybody, and indeed Martin has reinserted it. I once noticed, fortunately, when I had edited an out-of-date version of a page: I think Warren has a point. Ro Thorpe 19:00, 12 March 2008 (CDT)

I think Warren just needs to be more careful. =\ --Robert W King 19:44, 12 March 2008 (CDT)

We have plenty of time before the editorial council is replaced in April to figure out what to propose to trial. Perhaps we should put this proposal on hold for a week to get a fresh perspective? Warren Schudy 17:43, 12 March 2008 (CDT)

I guess I'm a fan of devolution and wouldn't object to different workgroups trying out different pilot schemes (one of which might be a template footnote declaring that Articles in XX workgroup are written collaboratively, and the policy of the workgroup is that no authors should be identified individually). I should perhaps explain why I have declared that I will not be identified as an author, and it's no slight at all on Citizendium articles. It's just that in my academic life authorship is understood as carrying responsibility for the integrity and quality of the whole article - authors writing in collaboration each are responsible for everything. It's no defence to say "I didn't write that bit." I wouldn't be able to say that for a Citizendium article without as Larry puts it, asserting some sense of ownership, which I very much do not want to do. So personally I would feel very uncomfortable - and I came here to write in partnership, in a different way, which I'm more than happy to do anonymously. But others have different needs, equally valid, and I've no wish at all to prevent those being fulfilled.Gareth Leng 05:04, 13 March 2008 (CDT)

I'm curious how that works with you with approved articles, Gareth. In other words, could your name appear on only those, i.e., ones that are locked? Stephen Ewen 09:53, 13 March 2008 (CDT)

Well, this is certainly true about collaborative academic publications. In practice, we compromise with other authors' views on issues where there is room for that, and stick to our guns on issues where we are clear that we support only one position. With none of my collaborative publications do I feel any reservations about the content: in fact, the main problems are caused by the low standards of British, American and other publishing houses which decide either to remove the proofing stage completely or to ignore my corrections. For example, I have one Oxford UP book chapter with a nonsensical half-inverted diagram and missing text-labels, "corrected" names of authors whereas the originals were correct, and grammatical "corrections" into semi-literate English.

The question I would pose, Gareth. is: "in what way does the collaboration mechanism of CZ impede you in reaching the same degree of shared responsibility as occurs with normal publications?" Do you argue less on CZ than you would in professional activities? This might be because on CZ you are writing on topics outside of your professional training...Martin Baldwin-Edwards 10:14, 13 March 2008 (CDT)

On CZ (as an author) I would contribute a bit (on something I know about or can find out about with relative ease) to an article much of which I may not know about; that seems fine to me. I don't feel the need (as an author) to insist on endorsing the final product as a whole and don't have the time to do what would be needed to do that. That's true for approved articles as well (in general, there might be exceptions).Gareth Leng 04:09, 17 March 2008 (CDT)

The way forward

I am hereby inviting someone--anyone--to join me as co-driver of this proposal. The co-driver would be responsible for spelling out, in as much detail as I have in the above explanation section, how the alternative template and rules for its use would work. It would be best if this co-driver could try to get various other people, who feel the current proposal does not go far enough, on the same page. The alternative should be explained in its own section such as "Complete explanation of alternative proposal."

Then we can establish a testing/pilot project period, which I think should last two months. Since it looks unlikely that there will be any agreement on how to choose the winner, I now think that we should not try to establish any standards of success in advance. We can collect our data and make our respective cases--if necessary--after the pilot project period is concluded. --Larry Sanger 15:01, 21 March 2008 (CDT)