User:B. Wolterding/asfd

Process is important. I feel that Wikipedia is in need of a new process, tentatively called Article series for discussion. Recent developments have made it clear, once again, that

What article series are
While subarticles are not supported on Wikipedia, it happens at many places that articles are organized in clusters, often hierarchically, around one common topic. This starts with small sets of articles which are built using summary style, but often goes far beyond that, leading to sets of 100 and more articles. These article series might be, for example,


 * An article about a city, plus one for each quarter or suburb, plus (possibly) one for each neighbourhood;
 * An article about a historical person, supported by several auxiliary articles which detail certain phases of his life and work,
 * An article about a musical group, one article for each of their released albums, plus articles on released singles and other songs;
 * An article about a TV soap opera, plus several articles and lists on strorylines and plot elements, plus more than 250 individual articles about recurring (and non-recurring) characters;
 * An article about a producer of consumer electronics, plus articles on their product families, plus articles about several individual products.

For the purpose of this essay, I will refer to "article series" as any set of articles around a common topic, anywhere between five and five hundred in number. Note however that the "common topic" is important: A selection of 100 articles about elementary schools in the U.S. might be very similar in nature, but it's not an article series. The set of all elementary schools in one school district, however, would be regarded as an article series.

Problems with article series
There are many good article series, often coordinated and maintained by a large number of editors or even a WikiProject. These are well-maintained, appropriate split into articles of reasonable size, and their content is relevant and in line with our policies and guidelines. However, not all article series are of this kind.

It can happen, for example, that an editor creates one article and several stubs relating to sub-topics, probably planning to extend them at a later time, or hoping that they will be extended by other editors. In cases, however, this never happens. This is particularly so when a topic has a natural hierarchy of detail, and it is a priori unclear to which level Wikipedia can or should reasonably cover it.

In other cases, editors create articles in a series because they mistake a natural hierarchy to be an appropriate split between articles. This may also be guided by bad examples within Wikipedia. (If this-and-that minor band has one article per song, why should my favourite combo not have the same?)

Other article series seem to grow in a more uncontrolled fashion, with various editors creating articles on a very low level of detail, sometimes off a list that somebody left with redlinks. Most prominently, this is seen for per-model articles on consumer electronic products.

In most of these cases, article series consist of a large number of stubs. Stubs are not bad as such, but a bad factorization of an article series does not really help the reader. Also, stub articles look so empty, right? Probably this is the reason why they are often "filled up" by inexperienced editors with inappropriate material, such as


 * repeated content ,
 * unencyclopaedic material, in particular extensive plot summaries,
 * statistical data, lists of technical features, track listings, directory-type listings,
 * trivia.

I should not conceal here that there are article series which do not have the above-mentioned stub problem, but which have rather grown for the very purpose of capturing large amounts of inappropriate material, specifically plot summaries of series of fiction (TV episodes and characters, usually).

Problems when cleaning up article series
While the above examples of article series are quite various, they have one thing in common: They are in need of cleanup. Cleanup does not mean that the series needs to be deleted; but more often than not, some kind of reduction is warranted, either in the number of articles, in their content, or both.

Wikipedia's workflows, much like its content, are organized by article. Thus, cleanup of article series is often incomplete: One article at a time gets tagged with notability questions, or merged, or proposed for deletion. Apparently this does not lead to consistent outcome, and is also somewhat inefficient. A coherent discussion about the entire series, and a consistent implementation, would be preferred. Still, one-by-one is currently our usual way to deal with such series.

Cleaning up problematic article series in this way may be a lot of effort. (That's not my main point here.) But moreover, cleaning up such article series can be controversial, as it would be with regular article deletions.

Why centralized discussion is warranted
Even if no controversy is at hand, ..

It has been pointed out to me that

Rather, I feel that controversial article series should be handled in the way that is common on Wikipedia for controversial matters: by centralized discussion. However, it seems that a proper process is missing at this time; and the AfD process, which is often used as a de-facto substitute, is not really suited for the problem.

Why AfD doesn't cut it
The most-used formal process on Wikipedia, and in fact the one which is de facto often used to handle the problems above in controversial cases, is the Articles for Deletion process (AfD). Experience shows, however, that AfD is not well suited for handling the problems of article series.


 * AfD is for deletion only.

The AfD process, as the name says it, is intended for article deletion. While there can be other outcomes of an AfD discussion than deletion (such as: keeping, merging, or redirecting the article), deletion policy requires that at least the nominator wants the article deleted; and "AfD is not for cleanup" is a phrase often heard in discussions.

On the other hand, article series hardly ever require deletion right away. They do require cleanup, even if this often means a reduction of content, or a reduction of the number of articles. However, it is not possible for the nominator to list an article with a merge proposal.


 * AfD discusses single articles.

The AfD process is designed to discuss one article at a time. While it is possible to include multiple articles in one nomination, such mass nominations often cause problems.

On the other hand, nominating the articles of a series individually for AfD does not reflect the similarity of articles, does not really contribute to a consistent solution, and seems to be a waste of effort.


 * AfD is too simplistic.

In extension of the issue above, AfD seems incapable of handling article series because these often require a more complex solution than a simple keep, delete, or merge. This solution would first have to be found, developing and evaluating several alternatives. With editors just voting keep or delete, AfD mass-nomination discussions are often not structured enough for such debates. More often than not, the AfD ends up in a "trainwreck", and is closed without result because no consensus could be found.

Requirements for a new process

 * It needs to be quite formal.

In fact, it needs to be more formal than AfD is today. Changes to article series have a lot of impact, and cause quite a bit of work.


 * It needs to be efficient.

In my point of view, there are a large number of article series on Wikipedia that would need cleanup, and where at the same time may cleanup may be controversial. Currently, if any of these are discussed at all, they are discussed on a per-article basis, which often increases workload at AfD. One of the goals of the new process would be to make these cleanup processes more efficient than they are today.


 * It needs to include more general options than deletion.

As noted above, the outcome of an article series discussion is hardly ever as simple as keep all or delete all. A target structure for the article series has to be found, that allows the content to be developed in the near future. Finding and discussing this solution, or multiple alternatives for it, should be an explicit part of the process.


 * It needs to have the involvement of experienced editors.

As with AfD, the intention of a centralized process (rather than individual discussion on the article's talk pages) is to bring controversial cases to community attention. In particular, the process should be used for article series that violate policies and guidelines on a large scale. It is therefore crucial that experienced editors, who are familiar with policies and guidelines, regularly attend the discussions.


 * It needs to provide technical assistance.

Dealing with article series "by hand" can be very time-consuming, and if it's only for posting a message to all article's talk pages. (This may be one of the reasons why few editors attempt a cleanup.) In order to simplify matters, some "regulars" at the article series process should be equipped with tools that facilitate the standard procedures, e.g. for tagging all articles of the series as "up for discussion".

Description of the process
The following describes a rough outline of how I would see the new Article series for discussion process. The discussion of an article series would comprise three steps:


 * Phase 1: Determining the scope of the series
 * Phase 2: Finding possible solutions
 * Phase 3: Deciding on one solution

Each step is aimed at finding consensus, which is determined by a moderator, an uninvolved experienced editor that oversees the process (much like the "closing admin" in AfD).

Nomination
Any editor can nominate article series for discussion. In nominating, he or she should


 * identify which article belong to the series,
 * make plausible that there is a problem with the article series, such as a violation of policy, that affects multiple articles,
 * sketch at least one possible solution.

Phase 1: Scope of the series
In a first step, editors discuss the the scope of the nomination. The goal is to identify a coherent series of articles that shares a common problem and can conveniently be discussed. To that end, the original nomination may be expanded (more articles included) or reduced.

The nomination can also be rejected entirely if


 * the nominated articles do not form a coherent series, even after reasonable expansion or reduction of the nomination,
 * the article series does in fact not have any common problem worth discussing,
 * the series had already been nominated recently (say within the last 6 months), and no substantial new arguments or facts have been indicated.

If no consensus on the scope of the series is found, the original nomination prevails.

A list of articles which form the article series to be discussed.
 * Outcome of this phase

Phase 2: Finding valid solutions
The purpose of this phase is to find possible solutions for the article series.

Every editor can propose a solution. These may be a mixture of e.g.


 * leaving the articles, or part of the articles, as they are,
 * merging articles into other articles of larger scope (existing or to be created), possibly abbreviating the articles in this step,
 * splitting existing articles into many,
 * relocating articles to other wikis,
 * deleting articles altogether, or replacing them with redirects.

Many different proposals can be made, but it needs to be shown that they are


 * allowable, i.e. they are in line with Wikipedia's policies and guidelines (with the policies, at a very minimum)
 * feasible, i.e. they can realistically be implemented.

It is discussed whether the proposals meet the above criteria. Proposals are discussed independent of each other, and consensus is determined for each proposal individually.

A list of possible solutions which are both allowable and feasible.
 * Outcome of this phase

Examples
Suppose that the article series under discussion consists of the article about the TV series Jelly Soap, a list of main characters, and 38 articles which contain plot summaries of the individual episodes of the show.


 * One proposal could be to leave the articles as is, entirely. This would however be in conflict with Wikipedia policy, WP:NOT, and should probably been rejected as not allowable.


 * Another proposal could be to merge the episode articles into a List of Jelly Soap episodes article (to be created), with one-paragraph plot summary each. While this is probably allowable, someone would need to volunteer for creating the List of Jelly Soap episodes, showing that this proposal is feasible.


 * A third proposal might be to transwiki the plot summary articles to the Jelly Soap Fan Wiki. For showing that this is feasible, it would need to be made clear that Jelly Soap Fan Wiki will accept the articles, any copyright/licensing issues have been settled, and that somebody volunteers for performing the transwiki process.

Phase 3: Deciding on one solution
If Phase 2 produces more than one valid proposal, the purpose of the third phase is to find which one of them is the preferred one for Wikipedia. Editors therefore discuss (in one discussion) which of the proposals should be implemented. They can, for example, take into account which solution comes nearer to Wikipedia guidelines, which one has the least impact on existing content, or which one is more likely to lead to high-quality articles in the long run. The goal is to find consensus on which solution will be implemented.

One solution that is to be implemented.
 * Outcome of this phase

Implementation
After the solution has been fixed, it needs to be implemented (i.e. the articles will be split, merged, removed, etc. as decided). In simple cases, this can be done by the moderator directly; otherwise it will be done by volunteers as discussed in phase 2.

Since many articles may be involved, the implementation may take some time; the moderator should however take care that the implementation is finished in a reasonable time frame (say one week). If necessary, a temporary solution can be used, such as moving articles to user space.

The result of the discussion should be posted, for future reference, on the talk page of the "main article" of the series (maybe also on each individual article). Large-scale changes should not be made to the series without a clear indication that Consensus has changes.

History of similar proposals
The need for an article series process is not so new. At least one similar proposal has been trying to reach similar goals: Some editors set up a process for reviewing TV series.