CZ Talk:Article mechanics

For those who are looking for things that were removed in bulk in this edit, like missing link subsection targets, that stuff's all in the extended version here. J. Noel Chiappa 12:38, 25 February 2008 (CST)

Bold titles and introductions

Perhaps Citizendium could cut down on the practice of starting articles with "An article title is a ...". Starting an article with a definition ("A dog is an animal that barks.") is fine, but it only makes sense for the kind of topic you would find in the dictionary. Somewhere along the line, Wikipedians got the idea that all articles must start with a definition, so now there are pages that begin like

A list of widgets is a Wikipedia page that lists different widgets. Below ...

or

The 2006 terrorist bombing of Timbuktu was a terrorist bombing in Timbuktu that took place in 2006 and caused ...

Sometimes an editor realizes how redundant this is, and changes it into something like

In 2006, terrorists detonated a bomb in Timbuktu ...

as if no word in the article title were allowed to escape. In short, this convention causes redundancy and often makes article introductions awkward, sometimes almost unreadable.

But there's more: the convention also leads to the introduction of neologisms. When every article must start with a definition, one must come up with a definite term for every topic, even when no standard name exists. To take the subject area I'm most familiar with, math, there are plenty of mathematical results that are perfectly worthy of inclusion in an encyclopedia but which are unfortunately "anonymous". Of course, some title is required for the article page, but this is not the same thing as endorsing a definition. If someone named Smith has proved a notable theorem on geometric progressions, it is perfectly fine to write an article with the title "Smith's theorem on geometric progressions". But it can be misleading to start the article with

In mathematics, Smith's theorem on geometric progressions says that ...

with that particular formatting. This is an inconsistency on Wikipedia since Wikipedia claims not to be prescriptive, but it is perhaps not a problem for Citizendium. Since Citizendium is expert-edited, it perhaps has the authority to be prescriptive about definitions. (Note, however, that I'm only talking about introducing names for obscure topics; definitions for controversial words like "terrorism" are a separate problem.) Fredrik Johansson 15:42, 6 February 2007 (CST)

Excellent points, Fredrik. Clearly, we must be careful about how we formulate this particular rule. Feel free to try your hand at it. --Larry Sanger 12:01, 8 February 2007 (CST)

I believe definitions are (very often) necessary, so I usually open my contributions with a defininition. I don't believe it must be a rule, however. The 'rule' (if we can call it a rule) should be the reader's needs. If a user of Citizendium looks for, say, 2006 terrorist bombing of Timbuktu, he does not need a definition. He's looking for (qualified) informations on a topic he already knows a little. Same for single species (e.g., the reader typing Amanita phalloides is hardly looking for a definition, he's most probably looking for an accurate description of the mushroom, e.g., because he feels a stomach pain...).

On the other hand, cohomprensive topics must start with a definition. Take fractal: unless you are a specialist, the first thing you need to know (and you look for) is: what is a fractal? Same for pseudoscience in my view.

And, should terrorism (or Intelligent Design, i.e., very controversial topics) start with a definition? I say, yes. For example, an accurate definition of terrorism from an authoritative source could influence international relationships (a little). Not only these definitions are necessary, they also require a lot of responsability from the editors.

Let me also say that this page is a good idea. Keep going! --Nereo Preto 03:56, 9 February 2007 (CST)

Remark. A related discussion takes place on Pseudoscience talk page. Should we move it here? --Alex Halicz (hello) 13:36, 8 February 2007 (CST)

Probably... --Larry Sanger 20:56, 8 February 2007 (CST)

User Anthony.Sebastian (Talk) 18:56, 22 February 2007 (CST) offers the following introductory paragraphs of articles for which he started the Introductions. Note the lack of formal definitions:

Systems biology

As an academic discipline, systems biology aims to explain, predict and control the properties, functions and behaviors of biological, or living, systems—compartmentalized assemblages of interrelated, dynamically interacting, coordinated and hierarchically organized components.

Life

Biologists use the term life to refer to the process(es) comprising the activity of living, and to the entities that embody that/those process(es)—complex adaptive systems. The question turns on what precisely characterizes the 'process(es) of living'

Potassium in Nutrition and Human Health

To maintain health, the diet of humans must contain potassium, in its ionic form (K⁺). In 2004-2006, the Institute of Medicine of the National Academies of Science and its Food and Nutrition Board recommended that adult humans consume 4700 milligrams (mg) of potassium per day, which, calculated from the atomic mass of potassium (39.1 mg per mmol), amounts to 120 millimoles (mmol) potassium per day: 4700 mg/39.1 mg/mmol =120 mmol. That intake of potassium exceeds estimates from recent surveys of average intakes by the general population.

Pseudoscience example

The following discussion took place on Pseudoscience talk page. To centralize our efforts it was moved here. Feel free to add to this.

Let's try to keep things in order. Defining 'pseudoscience' is related to the problem of demarcation, but is NOT the problem of demarcation itself. So I suppose there must be something in the definition which is more than "a thing which is not science".

In my view, it is common sense that all 'pseudosciences' pretend to be sciences. This enter some of the definitions we find in dictionaries. I suggest we define pseudoscience according to those definitions. Such definition should try to be objective (i.e., no reference to someone that believes something) and should include the concept that a pseudoscience 'looks like' or 'pretend to be' a science. All agree? Should I try to post a new definition? Ciao! --Nereo Preto 04:20, 7 February 2007 (CST)

Some time back the article began: The term pseudoscience which combines the Greek pseudo (false), and the Latin scientia (knowledge), appears to have been used first in 1843 by Magendie, who referred to phrenology as "a pseudo-science of the present day" [1] In 1844 it was used in the Northern Journal of Medicine to describe "That opposite kind of innovation which pronounces what has been recognized as a branch of science, to have been a pseudo-science, composed merely of so-called facts, connected together by misapprehensions under the disguise of principles". - this is from OED. Before then it did start with an abbreviated dictionary definition (I think....). We might be going round in circles. Some editors feel it's better not to start with a dictionary type definition, but get straight into the tone of the article. I think I started out with a dictionary type definition but saw why others preferred a different approach. I'd worry about the opening of the article later, as there are probably bigger issues about content to get right. Its easy to go round in circles about what is probably a point of style in the end. However, go ahead and post whatever you think, its always much easier to decide with an example rather than in theoryGareth Leng 11:21, 7 February 2007 (CST)

Thanks. I see your point, I don't want to go back to versions that were discarded already. I'll study the history page before posting. But not today, I'm going to spend many (expensive) hours at SEM. I anticipate I believe an article should indeed start with a definition. I believe the reader of 'pseudoscience' wants, first of all, know what 'pseudoscience' is, and then, maybe, spend other time learning more. See you soon! --Nereo Preto 02:20, 8 February 2007 (CST)

We can have it both ways. If we were to adopt a format that -at the very top of the page gave a "dictionary" definitions, or depending on the subject, the etymology of the word, followed by some kind of "break" and then the body of the article, a user would quickly learn where to focus tor information. Ideally the "top of the page" information would be in a smaller- or at least different- font. The break can be something like a triple horizontal line. Nancy Sculerati MD 09:41, 8 February 2007 (CST) addendum-another way to do it would to have a "top section" with a slightly different background color-like a light blue or yellow. Nancy Sculerati MD 10:09, 8 February 2007 (CST)

Doesn't table of contents make a sufficient break itself? We may put a relatively short definition before the first section. If we feel that dictionary-like essentials are not enough for a vast topic, we start the article with the first section entitled e.g. "Introduction" (this is at least what I'm trying to do in Fractal). Proportions may depend heavily on the topic. For example, in the present article the first two phrases would make a basic definition (that could be then slightly extended to include the first paragraph in an assertive way), the rest of the present "zeroth section" would make a good intro. Just a simple thought. --Alex Halicz (hello) 13:28, 8 February 2007 (CST)

PS. We are invited to discuss the issue on Citizendium Pilot talk:Article Mechanics|Article Mechanics talk page. Shouldn't it be copied (or at list linked) there? ..done

...After a while I thought I could actually do this to show what I mean on a concrete example (and I was curious how does it look and feel). Here it goes. Since I consider it only a proposal (and by no means the only solution), feel free to revert it (in addition to our usual open-for-modification model). --Alex Halicz (hello) 13:48, 8 February 2007 (CST)

So at the moment the Pseudoscience article represents an outline of a solution for formating the front-matter. Any comments? --Alex Halicz (hello) 02:38, 9 February 2007 (CST)

It's good, but I'm suggesting a convention that's a bit more compartmentalized Look again please, and comment. PseudoscienceNancy Sculerati MD 14:12, 9 February 2007 (CST)
(copied from Talk:Pseudoscience) I changed it to a format of the kind I am talking about. This is just an example. I took the old introduction, which discusses the word derivation, and put it in a top section set apart by italic text in a smaller font. The idea is that one's eye can skip over this, or- if looking for the word derivation, focus on it. I think this is a useful format for most articles. Nancy Sculerati MD 14:11, 9 February 2007 (CST)

I find "compartiments" idea interesting. Well, it separates dictionary-like info from a general lead. Just a technical note: perhaps it would be better if we could put the following three items in three boxes side by side: table of contents, "dictionary" definition and an image (if we have it). Now the "dictionary" part looks like another disclaimer, the lead (perhaps the most important thing on the first sight) gets shifted down while we have quite a bit of blank space next to table of contents. I do not know if I'm skilled enough to actually make the technical changes I'm proposing, but I'll try. --Alex Halicz (hello) 01:34, 10 February 2007 (CST)

...Here it goes. The three boxes in a row do not fit (not enough space). I tried 2 by 2 arrangement to preserve the proposed "skip or focus" idea. The question is whether it looks better than the previous version. Generally, I'd like to give some prominence to the "main lead", since this is what I think most readers look for. From this point of view, putting the "dictionary lead" _not_ above the "main lead" looks intresting. An alternative solution (more radical, though) would be to move the dictionary essentials to a footnote. Another possibility is to include the dictionary esssentials as the last paragraph of the lead (small font, why not). Any thoughts? --Alex Halicz (hello) 10:16, 10 February 2007 (CST)

Well, in Biology (you'd have to follow changes on the talk page going back to the first part of November,to see all the changes - we kept trying it different ways) I think it ended up as a foot note. But the "top space" text area has an advantage of being multi-purpose- it can serve as a spot for a definition, for an etymology, for disambiguation or all three. Take a look at Vertebral subluxation please, where disambigiation is a particular concern because of the content of the article. This is a bedrock notion (pretty much) in the "alternative medicine" of chiropractic which happens to use words also used by physicians with, as it happens, opposite views. We have been struggling with giving chiropractic the respect it deserves, without misleading a user into confusing a diagnosis of "subluxation of the vertebral jpoint" that they might receive from a neurosurgeon, say, as being described in Vertebral subluxation. This top text area seems to serve nicely. I think that the more I see it, the more natural it looks. By the way- vertebral subluxation is up for approval so nows your chance to comment! Nancy Sculerati MD 10:58, 10 February 2007 (CST)

Semantic info boxes generally

I am not opposed to the general direction you are taking this in, but I think we need to take a step back and think about some general principles and make sure this is indeed a good direction to move in. Perhaps it is my background in theoretical ethics, I am always thinking about whether policies can be generalized. The idea here is that there is something special, unusual, about what we might call semantic information, and because it is different from the bulk of the article, it can usefully be placed away from the main part of the article.

Somebody might say that it is possible to classify all different kinds of information--historical, critical, artistic, etc.--and that each deserves its own special box, say, each with a different distinctive color. But we can all agree that that would be silly. The question then is why semantic information is thought to be so importantly different that it should be given this different treatment. What's the reason? Well, I think it's something like this. When someone is trying to learn about a topic in general, all this business about the mere meanings and etymologies and histories of words is merely a distraction. If you put semantic info in the front (in the body of the article), then you defer the reader from seeing the more interesting stuff. If you put it anywhere else--except perhaps the very end--then you interrupt the flow of the article. But it would be better to put it at the very beginning, not the very end, because semantic information is introductory information, if any information is.

Philosophers since Quine and his contemporaries have debated whether a viable distinction can be drawn between merely semantic information and other ("synthetic") kinds of information. While I personally think such a distinction can be drawn (and I am probably in a minority of philosophers on the question), I think a similar point can be usefully applied here. In an encyclopedia article, the information considered merely semantic does tend to shade, by degrees as it were, into more "substantive" information. In the case of pseudoscience, we see already that the body of the article itself begins with a definition, and another definition appears but as part of a historical source quotation in the semantic information box. Furthermore, the opening paragraphs go on at some length to clarify the concept, which some might regard as semantic information.

Maybe you want to say that it's not just any old semantic information, but etymology and supporting historical quotations (a la the OED) that belong in the semantic information box. If so, then I would ask: is it always the case, or only usually, that we will want such information prised apart from the main article? That seems to depend upon another question: will this info usually be irrelevant to the substance of the article? But, really, to a certain extent all encyclopedia articles particularly on general topics are grab-bags of disconnected information. The reason I enjoyed the "Biology" article so much is that it had a coherent narrative that nicely brought together all that otherwise disconnected information. We should aim at that as much as possible. The real question, then, is whether etymology and supporting historical quotations (assuming we want them) will usually not find a sensible place within any unifying narrative available for article topics. If not, then the idea of semantic information boxes might be justified. Otherwise, the information should be worked into the article narrative. The current introduction of the philosophy article shows how semantic information can naturally be used to introduce a topic--have a look. I think earlier drafts of "pseudoscience" did this as well.

I'd say that if the etymology/origin of the word or phrase is interesting, then clearly it can be coherently and usefully worked into the article by using it to introduce the substance of the topic. The etymology of "pseudoscience" can be used this way. If the etymology is not so interesting or revealing (as in "Biology"), then simply include a brief parenthetical note: "from bios + logos" (that sort of thing).

It seems to me--though I don't feel strongly about this, by the way, despite the length of this comment!--it would be better to include information that can be included, in this way, rather than to prise it off.

Historical quotations illustrating the original uses of a word pose a special case. Since we are not attempting to reproduce the OED here in the encyclopedia project, there is no obvious obligation on us as encyclopedia article writers to produce such quotations. So I'd say: if such quotations are particularly helpful in explaining the meaning of a concept, then by all means, include them in the body of the article. Otherwise, omit them.

Of course, you might disagree with my premise that we are not attempting to reproduce anything like the OED here. You might think that we should have, uniformly, that sort of semantic information for every article (about a concept). Anyway...duty calls and I'll leave it there.

I'm not specifically asking you to do anything, just trying to make a contribution to the debate. --Larry Sanger 12:28, 10 February 2007 (CST)

P.S. Every new widget on a page adds clutter to a page. Simplicity and cleanness is valuable, hence widgets must pay their own way. --Larry Sanger 12:31, 10 February 2007 (CST)

comment

This page should have been completed before the site was thrown open; most of it still remains unfilled.

Dates

Is there a CZ standard for writing dates? I see these two formats used most frequenting: day month year (e.g. 27 September 1968) and month day, year (e.g. September 27, 1968). Also, the Wikipedia standard is to linkify dates and years (e.g. September 27, 1968) - is this a CZ standard, too? --steve802 13:46, 2 March 2007 (CST)

The WP standard is only to link some dates in certain cases where there is a value to doing so. However, many dates in WP are linked with no apparent reasoning. I think we should only link dates if there is a purpose to doing so. In most cases, I would struggle to think of a purpose for a date link though a few examples might be 9/11 or the years 1066 or 1945. In such cases the link would have to connect relevant articles. So for example "The Fitz Allan family came to England William The conqueror in 1066" would be appropriate because that person was actually involved in some of the major events of that year but "Irene Ducas was born in 1066 and die died February 19, 1123" would not, in my view, be a valid use of the link since an article discussing the events of 1066 would provide little (no) additional information about Irene Ducas since the notable events of here life happened after 1066. Derek Harkness 07:31, 2 April 2007 (CDT)

The way WP as the articles on dates set up is more of a list of major events that happened on that date. I am new to CZ and don't know how articles on dates are written, but if it is anything like CZ then I think dates shouldn't be linked unless the subject the date applies to is included in the article for that date. It seems I've kinda gone around in circles. If anyone can't follow I'll be glad to clear it up.Drew R. Smith 08:28, 15 June 2009 (UTC)

Should abstracts have a place in Citizendium? An argument challenging encyclopedia tradition.

An encyclopedia is not a collection of crib sheets; it is as much about writing good prose as about giving verifiable information. If the latter were the only aim one could arguably structure a computer based encylopedia as a hard fact name-and-number database, with a language friendly SQL query engine as its front end, and “articles” being “written” by completing data entry forms.

Nowhere is the conflict between good writing and transfer of information more apparent than in the lead section of articles written strictly according to the Wikipedia introduction style guide. It is not that the writers are necessarily incompetent, it is rather that following that give-all-the-info-in-six-sentences style almost without exception makes for poor reading – at the extreme resembling outlines of class notes for high school more than prose. Never the less, it has become a standard at Wikipedia – a “consensus” guide arrived at after extensive interaction between a mass of unknown persons of varied intellect, language skills and logical abilities. This bothers me, since a conclusion drawn from such input has validity. That validity lies in the qualification that the contributors share of being internet users of information, and they presumably know what they want. My perception is that they responded to the identified want by throwing out centuries of human style and bringing in study notes.

Was this the best response? I think not. I asked myself what other opinions about introductions or leads would be, specifically those expressed by persons who appreciate good writing for what it is, emphasising the joy of reading good prose while learning from it. So I have been going through recommendations from various universities, colleges, and training sites about the introduction to theses, term papers and essays. One thing these guides have in common is that not one – whether from departments of language, graphical arts or mathematical statistics - recommends that the lead or introduction to an information article be a self-sufficient summary of the article. The dominant idea is that the leading paragraphs introduce the thesis of the work. The thesis here is meant in a broad sense, not “this is a summary of all the facts”, but rather “this is the context of the subject that I am going to discuss”, the idea that gives unity, coherence and thrust to the article. An example of such a “thesis” would be, for a hypothetical History of China article: “The difficulty of understanding Chinese society: How a study of the history of China gives contextual meaning to the people, politics and economy of China as it is today”. From this flows the structure, factual content, conclusions, and comments included in the article.

A Pearson Education article notes that “Generally, the thesis statement will appear in the final paragraph of the general opening”, and a number of the university writing guides affirm that the thesis is usually left to the last part of the introduction. I interpret this as meaning that one does not jump in feet first, staking a claim and then fighting to keep it, but rather paints a scenario, and then leads the reader to where the thesis fits in. There is a certain deceptive gentleness about such a style which makes for easy reading, corresponding to the natural art of story-telling.

Contrasted with this intellectual and artistic goal of a well-written lead, there is the idea of the all-in-one abstract-summary-introduction-lead, where the lead can be taken as a self-sufficient description of the content of the article. That is a push more than a lead. In Wikipedia this idea is so pervasive that there has even been a suggestion that a stand-alone abridged version Wikipedia CD could be made up of leads to articles only. I have not come across any guide, other than the Wikipedia pages, where such a requirement is set or recommended.

The crucial point to remember is that this opinion on a way of doing things would not have evolved in a vacuum; it is a response to a need. The need is that the person looking for information on the internet should be able to read a reliable abstract of an article, to avoid wasting hours reading through unsuitable papers.

My thesis is that Citizendium would add value for the reader, as well as gain prestige as an innovator and leader, if it were to implement a policy of abstracts - as opposed to fact filled introductions - for its approved articles. I believe that the long-worked-at requirements for article leads that Wikipedia developed is the wrong response to the right perception of a modern need. I believe that Citizendium should introduce abstracts of longer approved articles – “longer” here meant by analogy with journals which require abstracts for articles, but not for “shorts” like letters to the editor or book reviews. This is a more sophisticated response to the need which those Wikipedians correctly identified.

A look through some of the articles being worked on shows that disagreement about what should or should not be in a lead is a frequent point of debate, taking up much space and time. This deflects effort which should be directed at the rest of the article. A benefit to authors of having a system of coherent abstracts, as opposed to having to formulate definition-laden, fact-repeating leads, is that they would be empowered to get on with the work of writing structured, coherent, informative prose for the articles. Arguments about what should be in the abstract could be kept to discussion pages, not hampering the article itself; it may be composed before the article, and later modified, or left to when the article is considered suitable for approval (when it should almost write itself).

There could be other spin-offs from the introduction of abstracts. The practice of writing abstracts of articles could replace the phenomenon of a stub, eliminating the stub altogether, to the extent that one could discuss the idea of having approved abstracts, as entities separate from approved articles. These would be something along the line of better-than-stub approved place holders, while authors get on with the job of polishing the articles on separate pages.

I believe that the use of article abstracts may well become standard practice to the field of internet encyclopedia publishing, and that Citizendium can lead this change.

Christo Muller (Talk) 13:51, 4 March 2007 (CST)

I believe this might reveal a great idea. Think of this: scientific journals ended up to this after centuries of experience in publishing, so there must be something very good in abstracts after all. I believe abstracts could make long articles much more readable.

Christo, why don't you start a thread in CZ forums? I suggest this idea must be discussed deeply, and I don't know how many of us have this page in their watchlist. --Nereo Preto 13:09, 6 March 2007 (CST)

Yet, should not the introduction serve largely as an abstract in an encyclopedia? Perhaps that convention has its very good reasons too. Stephen Ewen 23:44, 6 March 2007 (CST)

Unless you know exactly what you are going to write in a complete article, you can't write the abstract first- chronologically. That can only come for the organically grown article after it is finished-at least in rough draft. So-though I think it is a good idea, it's something more to add once an article is in an approved state. Often, we are arguing-the first paragraph is not an abstract- not because abstracts are not good-but because the first paragraph so often becomes a bone of contention in edits - with the ever dreaded subtractive rather than additive efforts of various authors. Also- for some articles the abstract is a really great idea- like say for a biography. But for the more essay like introductions to broad subjects-like Pain or Biology, abstracts may not be so helpful. My thoughts. Nancy Sculerati MD 00:22, 7 March 2007 (CST)

Agree on the first point. In fact, I was even thinking of editors involved in the writing of the abstract - if not in charge of it - once the article is nominated for approval. Some journals do not have a real abstract, but the first paragraph serves as abstract (Nature, as far as I remember, has this policy). This is typical of journals allowing short space for articles. I believe an "explicit" abstract or a cohomprensive first paragraph both do it. Yet the real point here is to make more readable long articles. Perhaps approved articles are already so structured, but would it be good to suggest such structure in some policy pages? --Nereo Preto 00:43, 7 March 2007 (CST)

Leads

As a result of a disagreement at God, I thought that I'd raise this here. Should articles (and especially lead sections) give (at least initial) definitions in terms of the most common usage among (native) English-speakers, or in terms of the most common usage universally? --Peter J. King  Talk  10:17, 20 March 2007 (CDT)

If the question of how a term should be defined is very controversial, then probably no single definition should be offered right up front (as god does not). But for purposes of determining what items might be offered up first as examples, or what definitions might be placed first in a list of definitions, I maintain that by initially following the usage most common among native English speakers--or whatever the relevant constituency is--we avoid the most conflict in the long run. We must create a widely generalizable principle here, because people will return to this page looking to settle all sorts of debates. If we make it a general rule to follow the usage of most native English speakers (for very general topics like "god"), given that they will make up the majority of our contributors and users, we will avoid most conflict in the long run. Furthermore, if this matter of "pride of place" is very generally understood, then it will come to be seen as perfectly neutral. The reason the Christian God is placed up front in god is simply that that's what most native English speakers understand by the word--no other reason.

And, just to forestall those who think we are discussing anything more serious or controversial, let me assure you that we are only talking about "what things are placed first." This does not mean we are deciding how whole articles are to be biased. This is just a general way to resolve a remarkably common source of conflict. --Larry Sanger 11:03, 20 March 2007 (CDT)

Narrative coherence and flow

I took the liberty to polish some of the wording in this section. Andrew A. Skolnick 12:34, 2 April 2007 (CDT)

Gareth, on your flow comments, I'm often unsure of the best way to handle time sequence on articles about individuals. There may not be a single best way. When an individual is especially known for specific achievements, which are mentioned in the introduction, reverse chronological order works well if those achievements are the most recent (e.g., James Jones). For a complex figure whose whole career (and even early life) is of note, such as Ho Chi Minh, birth-to-death order makes sense. For figures that might have done their most important work in midcareer but made lesser but important later contributions, there's no simple answer. Any suggestions?

Hayford, want to weigh in here? Howard C. Berkowitz 18:14, 27 January 2009 (UTC)

webcitation.org

I'd like to suggest CZ obtain membership with http://www.webcitation.org for reasons that are obvious. In the meantime, individual authors and editors can use it just the same. Stephen Ewen 15:20, 2 April 2007 (CDT)

Stephen, perhaps you (and others) could draft an addition to the relevant page(s) explaining the usefulness of the service. Basically, I think we should link in the first place to the live version, but also link to an archived version, as from webcitation.org--linked cleverly somehow, such as linked from the "Accessed <date>" bit. I think the policy generally, being one that is not really familiar to me or likely to be familiar to others, needs discussion on the Forums, though. --Larry Sanger 18:10, 8 April 2007 (CDT)

Will do. —Stephen Ewen 02:49, 9 April 2007 (CDT)

It is at http://forum.citizendium.org/index.php/topic,764.0.html —–Stephen Ewen 03:33, 9 April 2007 (CDT)

"Self-referential" templates

I'd created a {{fact}} template, similar to the one in use at Wikipedia, but I've just been informed that it was speedily deleted. I share the view that such templates are a Bad Thing if included in approved articles made public, but aren't they useful in articles before that stage? They let authors and editors (who are the only people reading the articles) that there's work to be done; relying on Talk-page discussion isn't such a good idea, I think, mainly because, not everyone looks at it as they're random-paging through Citizendium (I generally don't).

Incidentally, deleting the template page has just meant that hundreds of uses of the template are now red instead of blue, as they no longer point to the help page on giving citations. That was my purpose in adding the page in the first place, of course. --Peter J. King  Talk  06:59, 9 April 2007 (CDT)

I don't think we need a {{fact}} template. The template is widely missused on WP. It's often put on articles and never removed or re-edited. If you have a concern over a sentence, either research it yourself and find a source, or post on the talk page asking for clarification. Derek Harkness 07:29, 9 April 2007 (CDT)

The latter is covered by my point about reading habits, though. --Peter J. King  Talk  08:53, 9 April 2007 (CDT)

Peter--all articles, whether approved or not, are public. Perhaps you didn't know that. People can see the sausage being made. I don't think the template is ever useful, for the simple reason that a longer explanation of why one particular point needs to be supported by references will always be needed. Without a longer explanation, possible only on the talk page, use of this template is merely puzzling and argumentative to other users. Besides, it's generally a bad idea to intermix navel-gazing templates, of use only to contributors, with text, of use to everyone--not only because the presence of the templates is distracting to readers, but because it causes people to forget that we are not writing for ourselves, but for that broader readership. --Larry Sanger 08:47, 9 April 2007 (CDT)

Ah, another thing I didn't know, you're right. OK, scratch the point about private/public (pity, though; I'd thought that that was one of the advantages of Citizendium — that the reader saw the swan gliding across the water and could ignore the feet frantically paddling underneath).

The way that the citation rules are drafted at the moment seems to say that citations are only needed for certain sorts of material, so the addition of the template wouldn't be arbitrary (and an explanation could be given on the Talk page, once the reader's attention had been drawn to it there). --Peter J. King  Talk  08:53, 9 April 2007 (CDT)

new section

A new section is rarely needed for fewer than three paragraphs; but a single section could be suitable for considerably more than that.

OK, how 'bout this: a short paragraph at the end of the article clearly does not fit into the section that the eight paragraphs above it fit into. So I put in a new section heading, the last section being only that short paragraph, in order to separate it from the foregoing paragraphs. (I'll see if I can find a really good example. Michael Hardy 22:17, 26 April 2007 (CDT)

On quotes

Regarding lengthy quotes, this document states, "The exceptions will, perhaps, be in cases where texts themselves are the primary subject of an article. Even in this case, extended quotations are to be used sparingly and only with excellent justification."

I'd like to question the wisdom of this as an absolute rule. With some few articles, it is simply of overwhelmingly substantial benefit to quote more at length. Recently, Mein Kampf has been an issue with this, but there are others where summarizing and paraphrasing diminishes the type of article that could be. For example, I have considered writing an article History of the Mariana Islands. For its early history, this can only be done with full impact by quite extensively quoting the truly unique words of the colonizers. Another example--Roman Catholic views on X, with X being a significant issue. Quoting at some length the unique words of the varied popes across time would seem in order. I am sure there are some other exceptions.

I think this portion of the policy needs to be relaxed some to allow for more, albeit rare, editor discretion to include more extensive quotations. CZ ain't paper, and we can do it (Britannica's article on Mein Kampf is a mere 421 words and contains not one quote from the work). More extensive quoting, in some cases, can provide for some truly illuminating articles that would otherwise not be possible, I think. We should leverage that opportunity.

If not in the articles, then Article/Tutorial should allow for considerably more extensive quotations.

 —Stephen Ewen (Talk) 03:05, 14 July 2007 (CDT)

To revive a hoary suggestion, I think that Stephen is 100% correct on this issue. While lengthy quotations from secondary authors would tend to inhibit contribution, there is every kind of justification for paragraph-length quotations (and usually not longer) from primary sources. Instead of some arbitrary cap at one sentence, quotations should follow the "piece of string" rule-- long enough to do what they need to, but no longer. Brian P. Long 17:36, 13 April 2008 (CDT)

I agree. For an example of a wiki (albeit not an encyclopedia) article with extensive quoting, see: [1]. I wrote most of that, and it would have been impossible without the quotes. Sandy Harris 09:05, 22 October 2008 (UTC)

I suppose that rare and well-defined exceptions can be found to nearly any policy... --Larry Sanger 12:07, 22 October 2008 (UTC)

On spelling

Is there a standard on British or American spelling? Randall Bytwerk 10:41, 3 November 2007 (CDT)

There is--it needs to be added to this document. See CZ:Editorial Council Resolution 0005. --Larry Sanger 10:47, 3 November 2007 (CDT)

On Style recommendations

I'd like to propose that we recommend a superficial style recommendation that on a lede accompanied by an image, the TOC should be on the right and the image on the left if the image is "tall", and for a "wide" image the image should be on the right.

I think this improves the layout of the articles substantially.--Robert W King 14:22, 14 January 2008 (CST)

Draft rewrite

Forum discussions have alerted me to the thought that this article is too long, too wordy and repetitive. I have drafted a rewrite on Draft article content. Comments?Gareth Leng 07:13, 24 January 2008 (CST)

Moved to CZ:Article_Mechanics/Draft_rewrite. Stephen Ewen 11:32, 5 February 2008 (CST)

Why did you copy-paste rather than use the "move" command? I just nominated the redundant Draft article content for speedy-delete. Warren Schudy 17:43, 22 February 2008 (CST)

Citation relevance

I've been working on IPsec and an editor has been adding citations. Most of them are helpful, but I feel that some are not. For example, in IPsec#The_role_of_IPsec references 6 and 7 point to the RFCs defining other security systems, SSH and SSL/TLS. Those are useful references, and they certainly should be cited in the SSH and TLS articles. Also, the IPsec article should (and does) wikilink to those articles. However, I feel that those citations have no direct relevance to IPsec, so citing them in this article is just clutter.

I think the problem is more general than just this article, which is why I'm starting discussion here rather than the article talk page (though I will put a link there).

For one thing, something much like it has come up before; see Talk:Cipher. In one case there I changed an external citation to a wikilink and wrote a stub Venona for it to point to, copying the citation there. I did not realise we already had quite a good article VENONA, so I had the wrong target for the wikilink, Apart from that, I would still say that what I did was obviously correct, though opinions clearly differ on that. Is there a policy on when to cite external sources vs when to wikilink? Should we sometimes do both, as the IPsec article currently does for things like SSL and TLS?

In the other case, I removed a citation (of an excellent and standard reference, Knuth) because I thought it belonged in Random number but not in Cipher. This now seems to me a debatable decision; perhaps seminal works like Knuth are an exception that should be cited wherever they are relevant. not just in the article where they are most directly relevant with others wiklinking there.

However, I think the general policy should say external citations should be used only for things actually required to support the article's arguments or give readers links to more information directly relevant to the article topic. Anything else should use wikilinks.

This does not appear to be inconsistent with current policy:

the following categories of claims generally do need citation:

   * direct quotations
   * claims with unique sources (such as survey results, or the finding of a particular paper)
   * implausible-sounding but well-established claims
   * claims central to the article

However, current policy has nothing (or nothing I found, anyway) on the questions I'm raising. Sandy Harris 15:16, 22 October 2008 (UTC)

First, if there is a question of removing a reference that an author thought was appropriate, the starting point is to ask about it on the talk page. To me, it makes far more sense for me to add references to an article rather than take them away. Since the Wiki software here, or at least before the last change, was unhappy with forward references to citations, I have found it practically useful to put a full citation into an article when I think it may be needed later. Without asking, you have no way to know how an author intends to develop the article and how relevant that citation may be in a later part; the form you are looking at might be much like someone writing down research notes before writing a conventional paper.

In general, the CZ convention is not to remove text without explanation, and often to announce the intention to do so before actually making the change. Especially when there is a dispute, an Editor does have the authority to rule something doesn't belong in an article, ideally when it belongs somewhere else.

Another area to be very careful is when you think a particular technique, already in the article, is not used. When I see that, I may ask for examples of its use, but I won't just delete it unless, in my qualified opinion, it's flatly wrong. In the area of communications security, for example, there are techniques that indeed make no sense in civilian applications, but are widely used, and for operationally specific reasons, in military environments. While I don't think it's come up at CZ, I do remember a few times at WP when several contributors, all with enough output to be credible, make statements along the line of "it doesn't quite work that way, or there are other ways to do that, but I can't get into specifics". That might not even be for reasons of military secrecy; I can think of some techniques that are in a patent application in process or otherwise proprietary. Here, at least, my credentials are verifiable enough that if I say something like that, it has some credibility. There are times where I've tried to source something, checked with colleagues, and we agreed both that it was true, but it wasn't written down -- it was IETF or NANOG hallway discussion.

Sometimes, the reference may even be there as an aid to the author, to have a link convenient while writing the article. There are a number of times where I've done a link, used it for reference, and subsequently concluded that the topic was sufficiently elaborated by new text that the link was no longer needed.

Could you elaborate on what you mean by clutter? I think I take a middle-of-the-road position here. There was a former contributor that refused to believe anyone would ever wikilink, so his articles were both immensely long, but also would contain massive oversimplifications.

On the other hand, there will be times when someone needs to print an article, or to use it on a portable device with no Internet connectivity. I want that individual to be able to find the article useful, and still not replacing a book. In general, unless material in an article is discussing a specific protocol feature, I don't put protocol message formats into an article, without a specific reason to do so.

As far as RFCs, and I think this is generally reflected among the Computers Workgroup editors, the automatic "RFC" references has the problem that something linked alone doesn't show up in the references section. I believe when there is both an article on the subject and an RFC, both should be linked. There are times that someone looks at a topic and wants a more tutorial elaboration, which would be likely to be in a CZ article. There are other times where an experienced reader knows the information desired would be only at the level of the RFC. I see no good reason to force the latter to have to go to a CZ article and to the source documents from there.

With the VENONA/Venona example, I recognize that the search engine's case sensitivity is sometimes strange. Nevertheless, I again consider it appropriate to mention, on the talk page, that you see some topic is missing. Perhaps the talk page is not the place. Larry does not like what he considers an outline on an article page; I disagree, but it's his call and not that important a battle. Sandboxes have the problem that metadata and certain other links don't work there, but talk page drafts have the problem of overwhelming the page, and also having the text be open to comment when all the author is doing is checking links.

A very good rule, I think, and I will work on it more myself, is to keep some notes on the talk page. For example, in some of the DNSSEC and IPSec articles, I identify what I see as problem areas and where I'd especially like help. In the top-level DNS article, I've been asking the question of how much detail should be there rather than in subarticles. It's not unreasonable to put, on the talk page, what you are definitely planning to add; if someone else is interested, they can communicate and say they would like to work on that as well. It's frustrating to have been working on cleaning up something in an offline editor, and, just before inserting it, find the text rewritten around a different perspective. This is avoidable with communication.Howard C. Berkowitz 16:11, 22 October 2008 (UTC)

What I mean by clutter is external links, this that show up in the reference list, that are not directly relevant to the article's topic. For example, TLS, PGP and SSH are all used in the IPsec as examples for contrast, to help make it clearer what IPsec does and doesn't do, and VENONA is used in cipher as an example of a particular problem. They are all important topics and deserve their own articles with full citations. However, in the context of IPsec or Cipher, they are peripheral to the main discussion; by all means mention them and provide wikilinks, but there is no need to have the original NSA article cited in Cipher or the Open PGP RFC in IPsec.

Sometimes it is more complicated. For example, IPsec uses block ciphers like AES. The IPsec article therefore certainly needs wikilinks to the overview block cipher and to AES. It should also have external links to the RFCs that specify how AES is used in IPsec. However, I'd object to a link to the original AES design documents from the IPsec article. The RFCs will have that link, of course, and (when it is written) AES will too, so the reader who wants to follow up on the topic will find it. I would see it as unnecessary clutter in IPsec. Sandy Harris 17:17, 22 October 2008 (UTC)

I agree with some but disagree with others. For example, VENONA is of historical interest, and someone might navigate there from a military or history article. I've had enough tinfoil hat experience that I prefer, in things that have been heavily classified, to have both the article and the government link. NSA, in a number of its classification guides, will say "the fact of" capability has been declassified, but not what it did. Literally, they declassified "the fact of" the B-17 flying ELINT missions, but separately had to declassify that it was intercepting Würzberg or Freya radar. Wearing my historian hat, "the fact that" NSA will talk about VENONA is significant. So, yes, I think the NSA link belongs in cipher. Always remember that an article may seem primarily for one audience, but it is not implausible other disciplines will have reason to look at it.

AES internals don't belong in IPsec, not just because it's too detailed, but because IPsec, as an architecture, should be agnostic to the encryption algorithm. It is public information that with approved equipment, and Type I keys provided by NSA, you can run at least collateral TOP SECRET through it. Are there, however, NSA IPsec implementations that use some non-AES classified encryption algorithm? I certainly wouldn't bet against it; the documents are usually fairly careful about distinguishing between TS and TS/SCI/SAP. Does this speculation belong in IPsec? No, because it's only informed speculation.Howard C. Berkowitz 17:53, 22 October 2008 (UTC)

Stressing copiousness of wikilinks, not just relevance

The original of this document stressed the importance of copious wikilinks. I still want to stress it. Even if they make a lot of red links in articles. This is actually important; wikilinking is one of the essential reasons Wikipedia grew as fast as it did. I don't encourage Wikipedia's sometimes absurd linking practices, but neither do I think we should link only the very most important words. In a project, and using software, in which adding links is so easy, and have such beneficial results for readers and for writers, wikilinks "are your friend." --Larry Sanger 13:53, 24 January 2009 (UTC)

What about the question raised in the section just above, wikilinking vs citing sources? Sandy Harris 00:00, 10 February 2009 (UTC)

Also, some of the questions at Talk:Block_cipher#Questions_for_editors. In a general overview article such as block cipher, should wikilinks mostly point to other sections of the overview, e.g. #DES rather than separate articles where possible? Of course the overview section should wikilink to the separate article. When there is no separate article yet, should we create it as a redirect to a section of the overview, e.g. Feistel cipher?

I think there is some sort of hierarchy here. Articles need both external citations and wikilinks. In general, though, external citations should not be repeated excessively. Even a classic paper need only be cited once or twice. For example, Turing Machine cites the classic TM paper, and perhaps Alan Turing should as well, but the dozens of other places where it is relevant should only have wikilinks. Few readers need the primary source; our explanation should be enough. Similarly, an overview article needs only one or two links to a more specific article on some subtopic; any other references should be to the overview's section on the subtopic. That will be enough for many readers.

This is somehow related to the level of readership. Links to the primary sources need to be there for some readers, for editors, and for CZ's credibility. However, much of the time what the reader needs most is links to explanations of related concepts at roughly the same level as what he's reading. Taking Turing again, the article on the Halting problem needs a link to T's paper, but few readers will actually need the original. It also needs wikilinks to Turing Machine, Incompleteness Theorem and others. Many more readers will use those; they help someone understand the main article because the are at roughly the same level. Sandy Harris 00:59, 10 February 2009 (UTC)

Bibliography

In editing the page Category Theory > Bibliography, I read the suggestion "For formatting, consider using automated reference wikification." So I've made the wikified citation. Now how should it be used here? The instruction should be more specific. Thanks, Peter Lyall Easthope 15:04, 16 April 2009 (UTC)

I'll look into it. Chris Day 15:14, 16 April 2009 (UTC)

For the record, i just added the title of the book into that search box of the page found after using the hyperlink at the top of the bibliography page. It found the book using google scolar, then I clicked on a {{wikify}} link that was assiociated with the book. The code for the book in the {{citation}} tempalte was then spat out ready for copying and pasting into the bibliography page. I assume this is what Daniel Mietchen was expecting when he added that note at the top of the bibliography page. . Chris Day 15:26, 16 April 2009 (UTC)

Conversions

I've added a conversions section giving a quick description of conversion templates, and where to find them and their directions.Drew R. Smith 08:16, 15 June 2009 (UTC)

Lemmas and such

I only have time for a brief note at the moment, but let me observe one area of confusion: we have more than one "thing" that is either an article, or article-like. Several of them function in R-templates.

• Regular article: has, at least, a main page, metadata, and talk page. It should have a definition page. Since the Related Article bot is running, it will have a manually or automatically created Related Articles page.
• Lemma article: has a Definition page, which is the source of content; no metadata; and a main page that only contains the {{subpages}}. It may have Related Articles, Bibliography, etc.
• Redirect: has a main page, no metadata, and may have a Definition, Related Articles, etc.

In various ways, all three of the above function in R-templates.

• Disambiguation page: no metadata, no definition, deprecated in R-templates.

Howard C. Berkowitz 18:52, 3 August 2009 (UTC)

Definition?

In section "Opening Section" it is stated that:

The first paragraph usually begins with a definition of the topic, [... ]

and the sentence continues with:

e.g., Philosophy, both the field and the concept, is notoriously hard to define.

Thus, we have here the interesting definition of Philosophy (why the capital?) as "a field that is notoriously hard to define". This definition applies to maths, physics, history, geography, and not to forget, sociology as well. In other words, it is empty.

--Paul Wormer 06:09, 7 September 2009 (UTC)

Should key terms in definitions be linked as well?

The title says it all: Should we link terms in definitions? For example:

• Coffee Party [r]: Self-described independent grassroots political movement formed in reaction to the conservative Tea Party Movement. ^[e]

Should "Tea Party Movement" be linked as it is now? If this is the case, it should be mentioned on article mechanics page--right now, the current example ("Number") has no linked terms in its definition. Nick Bagnall 08:48, 21 March 2010 (UTC)

I think that style and usage, etc. will have to be reviewed by the EC (once it exists). As I see it, a definition can have links, but only relevant links that are recommendations to the reader. Thus I think that it would be wrong to link "mathematics" (why should a user go to a general page on mathematics when seeing this definition?), while it seems natural to link to "Tea Party Movement". --Peter Schmitt 11:22, 21 March 2010 (UTC)

Well, what about a definition like this one for The Manchurian Candidate: "Second and most famous novel by the American political novelist Richard Condon." Seems to me that the author should definitely be linked. Of course, I'm still not really sure just *why* we have definitions in the first place, beyond the fact that they satisfy some inner needs for various people who have come up with their concept and that, apparently, they serve some sort of mechanical purpose in the inner works of the the metawiki or whatever. For me personally, they're useless and I could live perfectly happily without them.Hayford Peirce 16:42, 21 March 2010 (UTC)

I think they're quite handy for the related articles subpage because they provide context (isn't that the next Write-a-thon theme?) without forcing readers to open new pages in their web browser (provided that those readers are just interested in a term's definition). Perhaps they're most useful as glossaries then? It'd be great to get feedback from people who aren't editors or authors. Nick Bagnall 16:52, 21 March 2010 (UTC)

Hayford, I agree with you: Richard Condon deserves a link in the definition.

It is mainly the Related Articles. It could also be used on the Main Page. Use in glossaries (just as in Catalogs) may lead to conflicts. Glossaries often need special definitions not equal to a general one. --Peter Schmitt 17:24, 21 March 2010 (UTC)

Two things come to mind:

1. If cluster main pages (and thus, by extension to CZ:Lemma articles, also CZ:Definitions) always contained bluelinks, navigation on the site would have no dead ends, inviting the readers to spend more time exploring the content. Bluelinks require a) inserting wikilinks into source pages, b) having content in target pages.
2. We do not have clear standards yet as to when a wikilink is appropriate from an article page. There is certainly no easy solution to this issue but perhaps a rough guideline could be that if a more general article (e.g. mathematics) is already linked from the definition of another one (e.g. number), then it would not be necessary to have wikilinks in number that are already wikilinked from mathematics. This definition of scope would not apply to examples listed in definitions (like counting, ordering and measuring in the current definition of number), though I would link them too in case people are interested in the examples rather than the general overview article.

For both points, having links in definitions would be essential. A related point is that some search engines currently pick up our disclaimer but if the definitions would be displayed on a cluster's main page instead of the disclaimer, a good definition may lure readers in, and if it contains wikilinks, we're back at point 1, so they can start navigating around right away. --Daniel Mietchen 19:55, 21 March 2010 (UTC)

I'm firmly in favour of links in definitions. There need to be more of them. Obviously one can have too many links, but I think that if the same text were to appear in the lede, it would be linked - and there is nothing particularly special about definitions and lemmas. –Tom Morris 20:26, 21 March 2010 (UTC)

Daniel, I am for carefully selected wikilinks, whether on pages, in Related Articles, and in definitions. They should be inserted where the user may profit from following the link. In my opinion, linking to mathematics from number is not such a link. A user going to a general -- top-level -- article in mathematics from number will not find anything useful there (while a link from mathematics to number is justified, of course). (With equal right one could -- but should not -- link to philosophy, psychology, language -- and perhaps even more,) Measuring will probably deal with measuring methods in science and thus need numbers, not help to explain them. ordering should point to order relation, also a page I would not recommend to consult in this context, and counting -- if such a page exists -- would have to direct back to number. --Peter Schmitt 00:00, 22 March 2010 (UTC)

See also previous discussion of link usage above, #Citation_relevance and the following section. Sandy Harris 01:21, 22 March 2010 (UTC)

We need to be careful to not over link. On a related articles page related terms will be clustered together (that's the point) giving many links to the same article right next to each other. Consider if United Kingdom was linked in the definition of all the home countries.

• Great Britain [r]: The largest part of the United Kingdom, comprising England, Scotland, Wales and islands immediately off their coasts. ^[e]
• Countries of the United Kingdom [r]: England, Northern Ireland, Scotland, and Wales; which form the sovereign state of the United Kingdom of Great Britain and Northern Ireland. ^[e]
  • Wales [r]: A country of the United Kingdom that historically was considered a principality; population about 3,000,000. ^[e]
  • England [r]: The largest and southernmost country in the United Kingdom, and location of the largest city and seat of government, London; population about 51,000,000. ^[e]
  • Scotland [r]: A country that forms the northernmost part of the United Kingdom; population about 5,200,000. ^[e]
  • Northern Ireland [r]: Part of the United Kingdom comprising six of the nine counties of the Irish province of Ulster; population about 1,800,000. ^[e]
• Government of the United Kingdom [r]: Constitutional government where executive authority notionally lies with the monarch but is exercised in practice by his ministers, and is the collective name for these ministers. ^[e]
• History of England [r]: Add brief definition or description

This is why I tend to not add links to definitions unless there is something unusual. Chris Day 03:36, 22 March 2010 (UTC)

That's a great point, Chris. I hope that the final decision regarding this issue--whether it's done here or by an EC--is eventually included in the Article mechanics page. Nick Bagnall 04:19, 22 March 2010 (UTC)

I agree. This is just the type of reasoning I wanted to suggest. Moreover, when the definition is displayed, there is also a link to the Related Articles subpage. --Peter Schmitt 12:32, 22 March 2010 (UTC)

Just a typo

The title of Section 7 should not be "=== Grammar, spelling, punctuation, and usage==". Boris Tsirelson 17:09, 17 November 2010 (UTC)