User:Ling.Nut3/TheSea

John William Waterhouse
Boreas

April 2009 saw the publication of a point/counterpoint pair of editorials about Wikipedia. L. Gordon Crovitz offers a vote of confidence in the Wall Street Journal ("Wikipedia's Old-Fashioned Revolution"), citing Wikipedia's "growing professionalism" and suggesting that it is "fast becoming the best" encyclopedia. Meanwhile, Oliver Kamm at firstpost.com takes a position at the other end of the teeter totter ("Knowledge by consensus is Wikipedia's downfall") by offering bracing criticism. Kamm writes that "Wikipedia is indeed comprehensible, and some of its articles effectively mimic the language of scholarly reference. The venture is junk, nonetheless". He rounds off this drubbing by saying "there is no reason to accept or value [Wikipedia], still less use it for its intended purpose." Only after all the torches have sputtered out and all the pitchforks are once again leaning against the toolshed wall will any villagers who read Kamm's editorial pick up on the fact that he never provides any real evidence to support his critical stance. Perhaps a bit more noticeable is the lack of any offer of meaningful advice or helpful input.

“ Effective writing – writing which is thoroughly researched, well organized, comprehensive but selective and prioritized, and above all is clear – is not a journey from point A to point B. ”

Kamm is correct in claiming that many articles – even among the bronze-starred featured articles, which are supposed to represent our best efforts – are just a mishmash of relevant terminology floating in factoid soup. He's right again to fault a project that over-emphasizes the (un)reliability of WP:CONSENSUS as a means of determining excellence. As often as not, consensus works to nullify any value contributed by Wikipedia's exceptionally sensible guidelines and policies on the necessity of employing verifiable and reliable sources, and on the virtues of maintaining a neutral point of view without conducting original research. If you're looking for examples of where the process has gone awry, look no further than Wikipedia's article on New Mexican English. This is an article that flouts WP:GNG, WP:RS, and just about any pretense of factual validity. Little wonder that Kamm finds our wares resistible.

“ Consensual soup – all too often, that is what we are in the business of creating. ”

Our faith in WP:CONSENSUS neglects to note that many editors are inadequately equipped, insufficiently disciplined or insufficiently objective for the task of creating excellent content. They are similarly impervious to the very real differences between content that is excellent and content that is merely adequately stocked with verifiable facts. By extension, they are unable to constitute a review process that rewards only the former while sending the latter back to the drawing board. Consensual soup – all too often, that is what we are in the business of creating. We take our lumps for it in editorials and in the offhand comments of university professors across the world, but those lumps are often well-deserved.

It's hard, then, to avoid the conclusion that Kamm has his finger on a truth about Wikipedia, but he may not have corralled the whole truth.

Crovitz too has noticed the existence of WP:V and WP:RS, as well as the significant influence they exert over articles. Rather than joining Kamm in completely discounting them, however, Crovitz places too much faith in their power to heal Wikipedia's wounds. Both views are wrong, in my opinion, and to find the truth one needs to look in a different direction.

“ At some point, people should be told: Magical thinking that involves plugging bits 'n pieces of WIAFA onto a page doesn't create an excellent article. ”

There is a deeper problem within Wikipedia's article creation and content review processes, and neither editorial points out the ghost at the banquet: flooding Wikipedia-space with do's and don'ts wrapped in the cloth of policies and guidelines cannot transform the great mass of Wikipedia editors into effective writers. The run-of-the-mill Wikipedia editor approaches WP:V, WP:RS and WP:WIAFA as items on a checklist, or perhaps lines in a recipe. Such editors are fairly quivering with an unspoken faith that if some minimum number of the right kinds of ingredients are present, the cake will come out Betty Crocker perfect. Unfortunately, our uncritical reliance on such guidelines fosters this approach. At some point, the spell needs to be broken, the penny needs to drop, and the folks at the water cooler to be told: much as jumping and then jumping again while still in the air can't make you fly, magical thinking that involves plugging bits 'n pieces of WIAFA onto a page doesn't create an excellent article. It isn't really even guaranteed to create a comprehensible article, though as I mentioned before, our various policies and guidelines do interact with WIAFA in ways that at least reduce the number of false tracks an article can go down.

Unfortunately, rather than slapping down the crap that results, the content review process makes it possible to reward such factoid soup output with Wikipedia's seal of approval, thus perpetuating the headlong rush to mediocrity. The FAC process is set up so that one well-placed Oppose can counteract a host of Supports, but it is also set up so that the outcome depends on whatever reviewers happen to show up (as opposed to designated reviewers of professional or near-professional quality). If only the fans show up to review that article, well then... we get what we get. This gaping hole in the quality control process is a direct offshoot of Wikipedia's attachment to consensus, and its volunteer makeup.

Effective reviewers are reflective readers.Perhaps a more fundamental flaw with Wikipedia's content review process is that too many editors work with the article they see before them, rather than duplicating the research. This is not only human nature, it could be argued that it is an extension of WP:AGF. However, factoid soup is a colloid that can easily withstand the pressure of reviews which do not delve into the relevant literature. Such reviews are based on three things: examining compliance with WP:MOS (always), verification of facts as presented (sometimes), and considering whether the references seem reputable (sometimes). However, those steps are necessary but not sufficient. It is not terribly difficult to conform to WP:MOS. Despite the keening sounds you may hear in the background, which are various editors moaning about MOS being an impenetrable moat, it really isn't that difficult to master the essentials. Moreover, reviewers will always be happy to point out where you efforts have MOS-related flaws. It is also very possible to string together verifiable factoids from reputable sources that nevertheless do not present a coherent narrative—in fact, that is the major underlying assertion of this essay. In my opinion, excellent content requires excellent reviewers, and excellent reviewers should not take the research for granted. Reviews should start from scratch, look at sources (most especially including recent and/or oft-cited sources not among the article's references), duplicate the research, and look for holes or fuzzy patches. Questions that must be asked: Do the sources raise questions that our article leaves unanswered? Do the sources present explanations that our article elides? In much the same way, the text as it exists cannot be accepted uncritically. What is the controlling idea of a given paragraph, and how does it relate to a larger narrative within the section? How well or how poorly does the paragraph support the narrative? How well or how poorly does the narrative reflect the concerns addressed by scholars? Most of all, when you have finished reading, what did you learn? Can you prove that you came away with a better understanding of the topic than you had before you read, by reciting the main points? Can you explain the relationship between these points and any larger issues? Or was the text just one reasonable-sounding fact after another strung together like white popcorn kernels on a string, each bit looking similar to the previous one, but none standing out in your mind? Did the text evaporate from your mind when your eyes left the page? If so, the writer has somehow failed to select key facts and present them in a salient manner, supporting them clearly, and clearly relating them to other aspects of the narrative. Which leads to the next point...

Effective reviewers abhor confirmation bias. This could be restated as: It is Noble to Oppose. It's true that every article that sucks should not be an FA. The converse, however, is false: it is not true that every article that does not suck should be an FA. Grammar and language usage is the first litmus test of suckiness; articles with serious or numerous mechanical problems get the boot fairly consistently. Next test: lack of MOS-compliancy. Third test: articles with mediocre quality references. But ah, here comes the problem. Once articles pass those three hurdles, a subtle form of confirmation bias tends to set in. Reviewers tend to nudge themselves toward a hypothesis, "Maybe this article really should be an FA." Then to confirm their hypothesis, they tend to work toward supporting it rather than looking for new problem areas. They should instead be looking for ways to prove their hypothesis wrong. Specifically, they should be reading the sources.

“ To write is to think, and to write well is to think well. ” — David McCullough

Effective writers are reflective readers. "To write is to think," David McCullough once said, "and to write well is to think well." Time and again I've seen articles fly through FAC because they were overstuffed with chronologically ordered factoids embedded within grammatical (and sometimes even well-written) sentences. All too seldom, however, do the editors seem to ... stop... and ask intelligent questions about the text they are placing atop the Google search results for global display. Our guidelines and policies are powerless to point out the fact that both research and writing require diligent mental effort and a continually, recursively reflective attitude. Time and again the editors must return to square one and ask themselves, "Is this all there is to the issue? Have I missed anything? Did I leave out or under-emphasize anything that cropped up again and again during my research? Have I overstated anything? Have I organized and presented the information in a way that my audience could intuitively grasp and clearly recall? Did I place myself in the reader's shoes, imagining the basic questions they might ask ("What is the relationship between wind speed and air pressure?), and writing to the questions? Did I scoop a few unprocessed factoid dumplings from various sources, plop them down onto the page and refer to the result as "summary style", or did I unpack them in a way that successfully transfers meaning from the page to the reader?"

The most important question by far is also the most often-omitted: "Why?" Why did Douglas MacArthur exempt the Emperor and all members of the imperial family implicated in war crimes? Why did the Russian and Austrians fail to press their advantage after the Battle of Kunersdorf, remaining in place rather than marching on to sack Berlin? Effective writing – writing which is thoroughly researched, well organized, comprehensive but selective and prioritized, and above all is clear – is not a journey from point A to point B. It is not even a two-stage process in which one set of content-providing editors charts a course, and a second set of copy editors removes the major stones and obstacles from the path. It is instead a prolonged period of round trips from A to B and back again, continually scouting the landscape via ongoing research, smoothing the path for others to trod, and shifting its course to make sure it offers an unobstructed view of the most important landmarks. More importantly, it involves a background process of interaction with the meaning of the text – a process that resembles a hyperactive ping-pong (sometimes quiet, sometimes very public) of role-switching from "editor" to "reflective reader".

Effective writers are made, not born. If Wikipedia wants to improve the quality of its articles, it must have the foresight to stop relying on the talent that we get, and instead start a self-sustaining process of mentoring the contributors that we have. Professional-quality writers, editors and researchers should rethink their decision to volunteer their work as editors, and consider volunteering to teach others their craft. The correct forum for this process, in my opinion, is Wikipedia:Good articles... Our other option would be to start hiring professional writers, and that of course runs far too sharply against the grain of Wikipedia's distinctive spirit.

I would say that writing and research are hard work. I would love to say that writing and research are hard work. Unfortunately, I strongly suspect that my idea of "hard work" is vastly different from many other editors' idea of what those words mean. The fundamental problem with Wikipedia – and it is a problem that our reliance on policies, guidelines and consensus only exacerbates – is that far too few Wikipedians understand the hard work involved in writing and researching well.