User:FreeKresge/Notability

This essay describes my views of the nature of notability, with two caveats. First, this essay represents my opinion and (unfortunately) does not represent official Wikipedia policies or guidelines. Second, the notability guideline described in this essay is the guideline as it stood when this essay was written. Like all guidelines, Wikipedia’s view of notability is in constant flux. The current guideline at the time someone reads this article may or may not match the guideline described below.

Two aspects of my background influence this essay. First, I have a Ph.D. in psychology, so I will use vocabulary used in psychological research such as “construct,” “operationalization,” “discriminant validity,” and “dichotomous variable.” I will try to explain what I am talking about when I first introduce the concepts.

Second, I used to teach writing courses (in psychology). I have learned that setting and enforcing high standards is essential to high quality. Students will be lazy and turn in sub-par work if I expect nothing more. However, most students can do high-quality work if I make it clear that I expect high quality work. I have also learned that many students seem to think that the more information the better even if the extra information is trivial. This is very seldom true. A paper with more substance is better than a paper with less substance. However, a paper with a set amount of substance and a lot of fluff is worse than a paper with an equal amount of substance and little fluff even if the latter paper is considerably shorter.

Why notability is important
Wikipedia has an ambition to become an encyclopedia equal to or even better than existing print encyclopedias like the Encyclopædia Britannica. To achieve that goal, Wikipedia needs to have high standards in many areas, including for inclusion. One of the standards that could make Wikipedia a high quality encyclopedia is a notability standard. However, this standard must be kept high in order to do so.

Standards for inclusion in Wikipedia are so pitifully weak that Wikipedia is becoming a joke, which is why some critics call it “Jimbo's Big Bag O'Trivia.” Although there are many good articles, I would find it very hard to refute that claim. There are a lot of pages on minor celebrities, minor web comics, obscure artists, etc. At best, the existence of such pages sends the message that Wikipedia is about the mundane. At worst, Wikipedia looks like a forum for self-promotion.

At the moment, only 0.2% of all articles meet or exceed Wikipedia’s standards for a good article. Thousands of editors spend a lot of time trying to make Wikipedia better. Unfortunately, much of this effort is wasted. Every hour that an editor spends improving a page on a trivial topic is an hour that is not spent improving a page on an important topic. For example, one of the few articles that not only meets Wikipedia’s standards for a good article but exceeds those standards is the article on spoo. Admittedly, Babylon 5 is one of my all-time favorite television shows, but I still believe that the editors who brought the spoo article to featured status could have better spent their time on more important topics. If articles on trivial topics do not exist or are merged into more important topics, then editors can focus on improving articles on more important topics.

Problems with current definitions
Currently, the main criterion for notability is coverage in multiple media sources. I have several problems with this definition that I will explain below.

Notability is poorly defined as a construct
A construct is an abstract concept that cannot be measured directly. Most of the topics that psychologists study such as aggression are constructs. The problem is that nobody has a good idea of what the construct entails until it is defined. For example, a psychologist may define aggression as the intentional affliction of harm against someone or something else.

Most psychological research is quantitative. That is, researchers use numbers to measure their constructs. In order to study aggression, a psychologist needs to somehow use numbers (or, at least, categories) to measure the level of aggression. The actual measure of the construct is the operationalization of the construct. For example, a psychologist might operationalize aggression as the number, length, and severity of electric shocks that a participant (falsely) believes that she or he is administering to another participant (who is really a confederate of the experimenter).

Part of my problem with the notability guideline is that there is no clear conceptual definition of what notability is beyond unhelpful dictionary definitions. I get the idea that notability is operationalized by the number of published works on the topic, however I do not see what this is supposed to measure. Without knowing what notability is supposed to be, I cannot judge if the operationalization actually measures notability.

Problems with discriminant validity
The concept of discriminant validity refers to the extent to which an operationalization of a construct measures just the intended construct itself and not other constructs. In this case, the operationalization of notability also measures verifiability (which may soon to be incorporated in attribution). The existence of multiple published works is needed to verify the accuracy of statements made in articles, so it is not clear what notability is beyond verifiability. If there is no difference, then it is not necessary to have a separate guideline on notability.

Notability is treated as a dichotomous variable
Many people treat notability as a dichotomous variable. That is, a topic is either notable or not notable without any consideration to the degree to which a topic is notable. I understand why people think this way. Notability comes up most frequently in deletion discussions in which a topic is either notable enough or not notable enough to be retained. In fact, the current notability guideline says that notability is relevant only in deciding whether a topic deserves an article, which all but invites editors to insert meaningless trivia.

However, different levels of notability can and should be used for editorial decisions other than whether or not to have an article on a given topic. For example, different bands may have the following levels of notability:

Level 1 Not notable enough to be mentioned on Wikipedia. For example, most garage bands would fall in this category. Level 2 Notable enough to be mentioned on Wikipedia, but not notable enough to have its own page. For example, a band known only for a single song that barely cracked the pop charts may be mentioned on a page of one-hit wonders but would not need its own page. Level 3 Notable enough to have its own page, but no more. For example, a band with a few minor hit singles. Level 4 Notable enough to have its own page as well as pages on key members, albums, and singles. For example, bands with several major hits or those who have been highly influential on other bands. Level 5 Notable enough to have of articles on all key members, many people affiliated with the band, all albums, most songs, etc. For example, a band with many hits that also greatly influenced most other bands that followed and had a strong impact on society at large. In other words, The Beatles.

Even this division is an oversimplification as there are many bands at levels between those listed above. Many decisions would be based on the band’s level of notability. For example, when writing an article about a level 3 band, the author needs to include everything relevant about the band. When writing an article about a level 5 band, the writer needs to understand that the article will be little more than a summary and a directory to sub-articles.

The above series of levels would change depending on the category in question.

Permanency is incompatible with current definition
Officially, notability is considered permanent. Once a topic becomes notable, it is notable forever. However, as Andy Warhol observed, “In the future, everyone will be world-famous for 15 minutes.” Although Warhol should not be taken literally, the fact that he is so widely quoted suggests general understanding that attention can be both widespread and fleeting. If notability is defined as media coverage, then notability is seldom permanent. If notability is permanent, then it is not simply media coverage at one point in time.

For example, in the summer of 2001, the media in the United States spent a lot of time covering a handful of shark attacks. On September 11, something else happened that distracted the media for a long time afterward. If media coverage equals notability, then the shark attacks during the summer of 2001 were notable through September 10, 2001. After that date, however, just about the only time the shark attacks received any attention was in commentaries about how such things seemed trivial in hindsight. In other words, the commentaries discussed how nonnotable the shark attacks were.

If notability is permanent, then the shark attacks were never notable despite the media attention. Right now, the victims and those close to them are the only people who care about the attacks as anything other than an example of the trivial things that people used to care about. If the shark attacks were notable at one time, then notability is not permanent as they are not notable now.

I am sure that a defender of the status quo would say that notability means that a topic has attracted attention at one time regardless of whether anyone cares about the topic today. For example, the shark attacks achieved the state of notability and will forever be considered to be notable even though hardly anyone cares about them any more. If true, Wikipedia will be full of articles on topics that people no longer care about.

What is notability?
My conceptual definition of notability is the extent to which people are currently interested in the topic and will continue to be interested in the topic in the foreseeable future even if the people have no direct connection to the topic. The extent of interest is a combination of the breadth of interest (i.e., the number of people who have an interest) and the depth of interest (i.e., how much they are interested). Personally, I would love to implement the 100-year test in which notability is defined as the chances that someone 100 years from now care about the topic even if she or he has no direct relationship to the topic. I would even settle for a 50-year test or a 25-year test. I understand that such a test runs into crystal ball problems because we really do not know what people will care about in the year 2107. However, the past can be a good predictor of the future.

There will be errors. For example, I once thumbed through a book covering films from the previous year, which just happened to be 1938. It had entire chapters to films that have since been forgotten and devoted all of three lines to Bringing Up Baby, which, today, just happens to be considered the greatest film of that year. However, these errors can be fixed by deleting the forgotten films and eventually adding Bringing Up Baby.

Determining cutoff points
Even though notability is a continuous variable, many decisions involve specific cutoff points. These points will be based on whether or not the notability of a given topic is greater than a given percentage of all topics within a given category. The cutoff will incorporate both the inherent notability of the category and the extent to which the topic is notable for a member of the category.

For example, suppose the article on Franklin Pierce is being considered at articles for deletion. By my personal estimation, he is the least notable President of the United States. Therefore, I would conclude that the notability of Franklin Pierce is greater than 0% of all Presidents of the United States. However, the inherent notability of the category, Presidents of the United States, is so high that all members are notable enough to have Wikipedia pages. Therefore, my vote would be strong keep.

On the other hand, suppose that the next entry at articles for deletion is a band whose main claim to fame comes from an album that barely cracked the top-100 sales chart for only one week. That is still better than what 98% of all bands have accomplished, so I would conclude that the notability of this band is greater than 98% of all bands. However, I might also conclude that there are so many bands that fewer than 1% deserve Wikipedia pages. Therefore, my vote would be delete.

Operationalizing notability
Even though the description above uses numbers, it is probably best thought of as still being a conceptual definition. Even if the Wikipedia community completely embraces my ideas, I doubt that we will see AFD debates like the following:

Professor John Doe

I am nominating Prof. Doe for deletion because he is more notable than only 78% of all physicists, and I believe that only the top 16% deserve to have Wikipedia pages. S.Hawking 17:76, 29 Feb 2007 (UTC)


 * Keep I agree that he is more notable than only 78% of all physicists, but I believe that the top 31% deserve pages. I.Newton 00:00, 30 Feb 2007 (UTC)


 * Delete I actually believe that he is more notable than 89% of all physicists, but I believe that only the top 8% deserve pages. A.Einstein 25:16, 31 Feb 2007 (UTC)

However, the concepts in the formula should carry over into notability decisions:


 * Notability is category specific. The criteria for determining the notability of a neologism are very different from the criteria for determining the notability of a politician.  For that matter, the criteria for determining the notability of a book are different depending on whether the book is intended for a general audience or an academic audience.


 * The likelihood that a random member of a category is notable enough to reach a certain cutoff point varies greatly depending on the category. A random Olympic Gold Medal winner is almost certainly far more notable than a random high school student is.


 * Decisions concerning notability are made based on how the topic in question compares to other topics within the specific category. A decision involving the notability of a given athlete involves comparing the athlete in question with other athletes.  In fact, the decision will probably be based on a comparison with other athletes in the particular sport in the particular league who played the same position at the same time that the athlete in question played.

It is nearly impossible to create a set of criteria that can be used for all situations. Therefore, a general page like Notability will probably be able to do little more than set very vague guidelines for more specific notability guidelines. A general page can give criteria such as winning major awards or being very influential on future work. However, a general page would not be able to name all the major awards in all the various categories.

The actual criteria that will be set in more specific pages such as Notability (music) or Notability (academics). These pages can further define vague criteria set in a general criteria page and can set specific criteria. For example, the music page could describe Grammy Awards and Mercury Prizes as examples of major awards. The academics page could cite Nobel Prizes and lifetime achievement awards from major professional organizations.

As mentioned before, notability is really a continuous variable, so any criteria listed on specific pages would simply indicate factors that contribute to notability rather than the criteria that make a member of the category notable. These pages could also provide examples that would meet various levels of notability, like in the levels of notability for bands listed above.

Testing criteria
A simple way to test the suitability of criteria is the following:


 * Find several experts in the specific category.


 * Ask each of them to create a complete (or near complete) list the members of the category who clearly meet or exceed a high level of notability (such as level 4 in the example above).


 * Ideally, the lists should have high overlap with each other and, if posted on Wikipedia, should be made up almost entirely of blue links.


 * If the lists contain more than a few red links, then the criteria are too lenient.


 * If the lists have little overlap with each other, then the criteria are too vague.


 * If the experts have a difficult time coming up with a complete list, then the criteria may be either too vague or too lenient or the category may be too broad.

Media coverage as a criterion
I still believe that media coverage on a topic is more a measure of verifiability (soon to be called attribution) than a measure of notability. However, there is enough overlap between the concepts of verifiability and notability to use published works as a criterion of notability as well, albeit more stringently than in the current guideline. I would make the following changes:


 * The “multiple” should be dropped from the criterion. Far too often, multiple has been interpreted as meaning “two.” By this standard, I would estimate that there are approximately 10,000,000,000 notable topics, and I would have a difficult time defending this estimate against a charge of being too conservative.


 * Instead, the approximate number of published works should depend on the quality of the works. There is a continuum from blogs, forum posts, self-published works, and MySpace pages to small and specialty publications to major news sources to scholarly works published in major peer-reviewed journals. Notability conferred by coverage in hundreds of blogs, forums, etc. might match coverage in only two or three articles in top journals.


 * Regardless of the type of work, the sources still have to be independent of each other and of the subject of the article in question. For example, coverage in 200 different posts in a single forum counts as coverage in one forum, not as coverage in 200 different sources. A single newspaper’s interest in a topic would not confer notability all by itself.


 * As much as possible, the coverage should describe how the subject is notable. For example, a concert review that simply describes the song list and comments about the sound quality does not contribute much to the notability of a band. An article that dismisses a band as being trivial might even diminish the notability of the band, although events after the article’s publication could easily make that assessment obsolete. An ideal article would describe the importance of the band’s work and how the band influenced others.


 * The published works should be spread over a reasonably large geographical area and over time. For example, a plane crash that receives news coverage only in the area of the crash (and perhaps where the plane took off and was scheduled to land) and only immediately after the crash is probably not notable, even if there was significant coverage at the time and place. A plane crash that received international coverage for years after the crash is almost certainly notable.

Conclusion
I understand that some people want to be nice and allow any topic to be in Wikipedia. However, having weak standards keeps Wikipedia from achieving its potential as an encyclopedia. One of the areas in which Wikipedia has weak standards is in notability. Wikipedia would be a better information resource if it clarified its definition of notability and tightened its standards.