User:Bilby/Notability RfC analysis

The following is an attempt to analysis the Request for Comment on a compromise for the current Notability requirements. As such, it is being conducted here in the interests of openness, but at this stage is not intended as anything other than a personal attempt to understand the findings. (I'm also aware that something of this length and complexity may be of no interest to anyone but me, however I hope that it will help to inform my own input into further debate, if nothing else). Naturally, if this is useful to anyone else I'd be very, very happy, and if anyone wishes to work on this or assist in any way they are, of course, more than welcome to – it wouldn't be Wikipedia otherwise.

It should be noted that partial analysis is next to useless, so until the full document is analyzed no binding conclusions should be drawn from any statements on this page.

Methodology
For the sake of this analysis a basic discourse analysis approach was employed. It was not intended to be a full, formal, analysis, but a simpler, semi-formal, "quick-and-dirty" approach to garner a rough understanding of the issues involved. (The emancipatory interest in this situation suggests that time is a factor). The use of a semi-formal approach, while not as ideal as a full, formal analysis, nevertheless offers significant advantages over an informal analysis. Specifically:


 * The reduction of the impact of personal bias on the study.
 * Increased rigour.

The methodology involves a number of steps:


 * 1) Anonymize the data set by removing all names (they are replaced with non-identifying numbers).
 * 2) Break down posts into separate speech acts.
 * 3) Categorize each speech act according to type (for example, description, condition, supportive argument).
 * 4) If the speech act constitutes a stance (for example, a condition to be met, or a claim offered to support the stance made in the post) it is further categorized using a faceted classification scheme.
 * 5) When all the posts have been analyzed, the resulting speech acts are grouped based on their classifications.
 * 6) Finally, each grouping/classification is examined to gain a picture of the overall argument presented.

For example, the post:

"Support, I think that the notability of the parent article should be enough to qualify its spin-offs. I think that if the notability requirement becomes more stringent on dependent pages, they will be folded back into the parent article, leading to some very large articles. I like that the articles can be parceled out into manageable units."

provides four speech acts:


 * 1) Support
 * 2) I think that the notability of the parent article should be enough to qualify its spin-offs.
 * 3) I think that if the notability requirement becomes more stringent on dependent pages, they will be folded back into the parent article, leading to some very large articles.
 * 4) I like that the articles can be parceled out into manageable units.

The first is a descriptive statement of the stance being taken, and does not require further analysis. The second is a proposition, and is categorized as "notability stems from parent article". The third is another proposition, and is classified as concerning "article length". The fourth proposition is then classified as concerning "organization of content".

The final analysis phase is guided by four principles: all viewpoints should be respected; consensus is not about numbers; Wikipedia is about collaboration; and the intent of the analysis is to provide practical outcomes to help Wikipedia. The first principle is expressed by not evaluating the worth of any stance during initial analysis: each argument is regarded as worth examining and is expressed accordingly. This has the side effect of dramatically increasing the length of the analysis, but later stages should allow the different arguments to be merged into a tighter conclusion. The second is managed by focusing on the arguments, not numbers: numbers may be mentioned where appropriate, but this is not to involve a detailed breakdown of figures. The point here is to gain an appreciation for the arguments being expressed, not the numbers of people expressing them. In terms of the third principle – collaboration – anyone should be welcome to be involved in the process, and all of the data (raw, partially analyzed, or otherwise) will be gladly provided on request. Finally, the fourth principle encourages analysis to focus on ways forward, rather than merely being descriptive of how things are.

Proposal A.1: Every spin-out is notable
 Proposal: A spin-out article is treated as a section of its parent article. If a parent article is supported by reliable third-party sources, then its sub-articles do not need reliable third-party sources to qualify for inclusion. A sub-article is notable when it extends one section of a notable parent article.

Rationale: It is not desirable to delete sub articles with a lack of appropriate sources. It makes more sense to treat those articles as extended components of their parent articles. Splitting content from an article into sub-articles is a practice recommended by the recommended length of articles and summary style approach. By treating sub-articles as though they were sections in the larger article, this would allow editors to write detailed articles on specialized topics.

Summary

 * Some editors are opposed to the principle of notability.
 * Sub-articles can be warranted based on the notability of the parent article.
 * Creating sub-articles will improve the organisation of article content, resulting in shorter and more readable articles.
 * Sub-articles will allow Wikipedia to provide more comprehensive coverage of topics.
 * Sub-articles are not "indiscriminate collections of information".
 * Sub-articles may encourage editors to work on more general topics, in order to provide justification for more obscure ones.
 * Deleting articles not only removes possibly valid content, but also takes time. The proposal should help both problems.

Conditions


 * Reliable sources are still required.
 * Sub-articles should be clearly connected with the parent.
 * Content forks must be avoided.
 * It was suggested that a word count for plot summaries be provided, and that sub-articles are barred from being Featured or Good articles unless they establish their own notability.

Detail
59 users supported this proposal, of whom one was an IP address and one did not sign the post. Of those 59 users, 9 expressed "strong" support, while 16 supported either tentatively or offered additional conditions to their support.

A number expressed their support through stating their opposition to notability as a matter of principle. As they believed that notability was either not a concern or, in one case, that it was actively harming Wikipedia, the proposal's removal of notability requirements for sub-articles was a good thing. Arguments for this perspective included that notability requirements were harming the depth of coverage, and that they failed to reflect that actual usage of Wikipedia, as expressing in the page statistics. Of those who did not directly oppose notability, the belief that sub-articles gain their notability from the parent (as expressed in the proposal) had support. Interestingly, while one user argues that this was distinct from the concept of inherited notability, three supporters disagreed, but were comfortable with the idea of inherited notability in this situation.

Some users (five) expressed a possibly similar stance through the claim that "Wikipedia is not paper". In two cases no further explanation was offered, while one user expressed the belief that Wikipedia was effectively a "limitless encyclopedia of all knowledge", and presumably could (and should?) afford the space. However, the other two proponents expanded their reasoning and took a different line, arguing that the limitations placed on paper-based encyclopedias need not be the guiding principles on Wikipedia, thus allowing for different organisational structures to be developed for the information contained within. This version of the argument was thus not based on "Wikipedia has unlimited space", and one of the two specifically responded to the counter-claim that "Wikipedia is not an indiscriminate collection of information" by arguing that the information being discussed for potential sub articles was not indiscriminate, and therefore not open to this problem.

Indeed, the need to improve organisational structure appeared frequently (17 times) in the responses, either as an issue in and of itself, or in regard either to the readability or length of individual articles. Proponents argued that long articles are difficult to read, and thus breaking them into smaller parts will increase the usability of the articles. One proponent stated that there was a bias against short articles, and that short articles would provide a better solution than the current "redirect to sections" methodology. Along similar lines, one user expressed the concern that all relevant information for articles should be kept in Wikipedia, rather than requiring users to look into the various Wikas.

In terms of raw content, some users expressed a desire to have more detailed or comprehensive articles in Wikipedia, and believed that sub-articles would help with this. Other than allowing for more information to be added (with one person comparing sub-articles to appendices), one user argued that this proposal might make it easier to keep articles in which the only references stem from alternative or new media.

WP:AfD was raised in several of the support arguments. Three basic issues were expressed. The first is that the existing model makes it relatively easy to delete spin-off articles, even though the information contained within was considered valid when it was part of the original article from which it came. A very similar argument was raised by another user, who felt that obscure and difficult to source topics may be able to gain protection by being treated as a sub-article. This, it was argued, would encourage users to create general articles with which to provide additional support for the obscure topics. The third issue raised was that AfD debates take up considerable time, and that this time could be better spent writing articles. The proposal would free sub-articles from AfD debates, so presumably passing the proposal would increase the amount of time editors have available to work on other areas of the project.

Other expressions of support were based on calls to "common sense", that there was no functional difference between having content in an article and having the content in a sub-article, and that it was a suitable approach for "uncontroversial" topics. Only two users argued solely "per user X", and only one chose not to provide a rationale.

As mentioned, a number of users provided support, but with qualifiers. The most common of these was the ongoing need for reliable sources: it had to be ensured that the sub-articles would rely on reliable sources, even if they were insufficient to establish notability. There was also the belief that sub-articles should show their connection to the parent article in some way, so that it is clear that they are not stand-alone articles, and that there is justification for creating the sub-article. In terms of content, it was desired by at least one user that more detail will be provided on how content forks will be avoided, and another made it clear that anything in a sub-article would be appropriate in the main article. Original research should also be explicitly avoided, and there was a proposal that a word limit be provided for plot summaries. Finally, one user expressed a concern about the overuse of primary sources, leading to a risk of sub-articles of questionable value.

More generally, it was made clear that sub-articles should still be candidates for both merging and AfD, the latter in order to avoid "fanwankery", (although this stance is likely be in opposition to other supporters, who had expressed a desire for teh proposal in order to free some articles from AfD). In addition, it was suggested that sub-articles should not be candidates for Featured and Good Article status unless they establish their own notability, and that the criteria for sub-articles should not be developed at an encyclopedia level, but should only be handled at the level of the project.