1. Introduction

How much credit should be given to the authors of comments or the originators of ideas in wikis, and how should this be done? It seems to me the current system is fairly anonymous; at the creation of a page perhaps the content is signed by an author, but after awhile, as the page gets refactored, it becomes impractical to note who contributed exactly what, and so personal attributions are erased. I feel this is a good thing in general; however, it would be nice if it were possible for another mode of attribution, one in which authors could make contributions and get credit for them.

There are at least two different sorts of AuthorshipCredit which may be useful:

Which are?

1.1. Goal I: General name recognition

1.2. Goal II: Recognition tied to specific contributions

In a GiftWiki, one would expect goal II to predominate; in an AcademicWiki, one would have both goals. In an academic setting, one has the additional problem of giving credit for being generally helpful in the past to people even if they no longer help out as much. Also, one should note that you may want name recognitions for having great ideas, or you may also want name recognition for being generally helpful (editing, organizing, and teaching others).

General issues that affect many of the proposed solutions are:

you don't want to cause arguments
you want to reward people even if they are so humble as to not publically assert their own hard work.
you don't want any systems that are particularly prone to malicious users or small cliques gaming the system (AvoidTemptation? ?)

some comments specific to the academic case have been moved to AcademicWiki.

2. Solutions

2.1. Contributors Tag (both goals)

General name recognition of 'good' authors or editors may be desired.

The "Contributors:" tag below entries used on some MeatBall pages is one way to achieve this. One problem with this solution is that in a professional academic setting it seems that disputes would develop over people adding their names to the "contributors" line without having contributed "enough".

Another problem is that this solution does not lend itself to a personal choice by each author as to whether they want to get some sort of credit for an idea; for instance, on MeatballWiki (and maybe WikiWiki also) I would expect that sometimes "Contributors" tags are erased during refactoring. A wiki community would have to develop strong norms about the retention of "Contributors" tags in order for these tags to really fulfill their function. It would be nice to have a TechnologySolution that would allow some authors to retain credit for their contributions even if there were others in the community who didn't want to bother about such things.

: Conversations would be easier to track if subtle font colors, for example, were used to automatically color code authorship, of, say a single-authored sentence. And a minimum-fontsize name tag could be automatically inserted. (Thus, subliminal learning of "voices," augmented by the instant ability to say "who is this guy?"
: This doesn't solve the problem of refactoring/defacing, (which is a feature as well as a bug) but the Complementing Weblog fills that gap. --JonSchull

2.2. Side column listing editors (goal I)

One idea for the academic context could be for the Wiki to be displayed with a column to the right of the text that lists the author of each block of text in a small font. Right now this, combined with a Contributors tag, is one of my favorite solutions; it puts more emphasis on good editors that any visitor will notice. -- BayleShanks

2.3. On-demand Verification (goal II)

Instead of constantly maintaining crediting information, only supply AuthorshipCredit for certain individual contributions when demanded. For instance, in a GiftWiki, it may not be important to know who the contributors are most of the time, but at the point in which credit/payment for a good contribution is made, the author of that particular part would have to be known. It seems that this is a much easier problem than situation (1), above.

2.4. The Wiki tells you who contributed if you ask (probably only goal II)

A feature in the wiki software that allows you to paste text into a box and submit; the result is a list of people who contributed to that section of the current page (you could click on each person's name to get the diffs of their contributions to that section)

I don't think this is a sufficient solution to the name recognition quandary (goal I). Most readers would not have the time to use that function on everything they read, so the "contributors" line would have the most impact on their minds and hence on an author's standing.

-- BayleShanks

In the case of really discrete contributions, like songs in MathSongsWiki, contributions could be signed. disputes would be infrequent as one would have to maliciously change the tag to cause trouble. When making a contribution of the type for which payment may be expected, the author would press some special button to notify the wiki to remember this diff and who the author was forever. This way all diffs would not have to be preserved indefinitely. Maybe instead of actually saving anything, each wiki could have a PGP key, and the author would get sent an email note from the wiki containing their special contribution that says that they are the author; the email would be signed by the wiki. This sort of approach doesn't violate ForgiveAndForget. -- BayleShanks

2.5. AnnotationWiki (both goals)

Another approach to identify the authors is to use WebAnnotation [SideWiki annotating] Wiki pages. This appears to me a valuable tool for contributions to and from people external to a Wiki community. In an intra Wiki community, taking and giving can be balanced by information exchange, whereas in the wider context there must be some more convertible currency. Take a community within MathSongsWiki : if everybody in it would only exchange selfcomposed music with each other (or some nifty things round about it), they would starve to death. External people might rip off songs, place them on CD and have a nice living. Imagine a very prolific author/composer, that does 99% of the work, another bad Wikizen makes some minor changes and claims authorship. -- FridemarPache

ClearCase?, although terrible, has a feature that gives you the history of a file to the line. With that, you can find the first and last authors to edit a supposed line. You could use something like that if you really wanted to, but I think that would discourage people from editing other people's works, and the GiftWiki moreso. The GiftWiki essentially accentuates the importance of authorship, and the per line history determines authorship. The history violates ForgiveAndForget as an extra added bonus.

By the way, my hunch is that CritOrg has no legal right to modify the page database of a wiki, say, this one. I suppose I could exert my copyright over the PageDatabase to prevent them from creating derivitive works through "annotations". External "contributions" would be outside the protection of FairUse because they would essentially be creating new works with the entire contents of this site embedded. Consequently, I would guess that an explicit exemption by a wiki owner would probably have to be made for this usage of his or her site. -- SunirShah

Good point. Sunir, as far as I know, CritOrg server only stores:

the links to Webpages and special searchstrings for pointing to finegrained targets within the pages
(optionally) the original annotation texts, making fair use of the underlying primary material (that is not stored, but only pointed to). Another option is to use an arbitrary Webpage (e.g. a Wikipage) as the text, that annotates.

As a link database they appear to me as a userdriven link-resource, more humble than the [OpenText] project, the people that built the Yahoo engine. -- FridemarPache

Is CritOrg secure enough? i would be worried about people attaching bogus claims of ownership. i think there would have to be some sort of communication between the GiftWiki's software and CritOrg to verify ownership at the time of content creation. -- BayleShanks

2.6. Complementing WebLog (goal II)

A personally controlled (non-volatile) Weblog as the complement to the communal and plastic wiki.

: Why just last night I [blogged], "...there is a not-just-self-indulgent complementarity between wikis and weblogs. A Weblog is an archival chronicle of development. A Wiki is an instance of a developing system. ..."

: This lets me (and when the spider comes, Google) "sign" and date-stamp a contribution, provide a person-o-centric (rather than wiki-centric) context and move on. The [WebSeitz website] incorporates something like this, but there are perhaps advantages to having the data on two separate systems, especially given how easy it is to cut and paste from one browser into another.

: Well, I guess I'll copy this nugget to my weblog now...
: --JonSchull

2.7. Centralized Recognition (goal I)

Some sort of centralized, formal (technological or social) system for giving credit to good editors (credit not for original ideas, but for helping out). For example, a computer-maintained page of frequent editors (perhaps with some sort of wacky meta-moderation system), or a community-maintained page of really helpful people. Or a formal system for electing officers of the wiki, officership being a way of recognizing helpful people.

With either a community-maintained page or election of officers, the community would have to have a habit of nominating people; otherwise people might be too humble to put themselves on the list.

I feel that near-unanimous community approval must not be a precondition for getting recognized, at least in an AcademicWiki. It seems to me that many academics have at least a few enemies who would be able to block such a process.

2.8. Less Control: Save Now, Prove Later (goal I)

In terms of reputation, I think an ordinary Wiki is good enough. Without any theoretical background, using Meatball and the Emacs Wiki as a empirical source, it seems to be true: People "know" who deserves a good reputation. It reflects in edits, in RecentChanges, in comments on the homepage. If you really want to proove your reputation, all you need to do is keep all the past versions of a page. Then, you could build tools that help you analyse the page database, in case of dispute. For example, you could claim in your resumé that you helped build the Neuro Wiki in the early years, and use one of your tools to prove it. For example check all the edits of a given year, and see how many of those where yours. Or see what percentage of the text now remaining is yours. Etc. The point is that you can build these tools later, that all you have to do now is use your UserName when editing, prevent others from using your UserName (doable with SoftSecurity), and never expire pages (simple technical solution). You might even decide to change the way old versions are kept. RecentChanges might continue to use something akin to KeptPages (ForgiveAndForget in everyday use). Only special purpose tools built into the script would be able to access all the pages (your reputation tools). Then you can keep the spirit of KeptPages, and still "prove" things. The algorithms used, of course, might be a disputable. But that is not a problem, since you have all versions available, you can change the analysis later on. All of this without any logins, text tagging, etc. And efficiency doesn't matter too much, because these tools get used in the case of dispute only. -- AlexSchroeder

3. More Discussion

I think we have found a number of ways to satisfy Goal II (getting credit for a specific contribution); but Goal I is not done yet (gaining name recognition). Of course even on goal II there is plenty of room for discussion. -- BayleShanks

Someone making a post who doesn't desire credit is an AnonymousDonor.

Discussion

Possibly related: The most popular Free Software license - the GNU GPL - does not require attribution. Of course the CC-by license explicitly does. -- Patrick Anderson