Wikipedia talk:Writing better articles/Define and describe

Supporters of the "define and describe" rule
Tim Shell, Larry Sanger, LDC, Linus Tolke, drj, Mike Dill, Koyaanis Qatsi, GWO, tbc, AxelBoldt, Enchanter, Rotem Dan, Bensaccount

Opponents
user:24

Martin 00:19, 30 Mar 2004 (UTC) - describe, don't define.

Commentary on the "define and describe rule"
24 - in my experience people learn jargon far better from good examples, good in-context use. This would be a better rule: always include an "i.e." ("that is", restating the thing in different words), always include an "e.g." ("for example", giving a more specific case).

Tim, "to make finite" as a definition of "definition" is surely totally inadequate. That might (I suppose, I don't know & don't care) be the etymology of the word, but the word itself doesn't mean to make something finite, unless we're speaking metaphorically, which we shouldn't do in definitions. I mean, what does it mean to make a word or a concept or a thing (all three are sometimes said to be the subjects of definition) "finite"?

See fallacies of definition, last item, "Obscurity." --Larry

A definition establishes the boundaries of the concept being defined. Finite means having limits, and to define means to establish those limits. If a concept is not finite, if you don't know where it ends or where it begins, then you have not arrived at an adequate definition. - TS--

To LMS: Why are we arguing about a criterion for a definition when we know what it is by common sense...:-).

RI, good point. Nonetheless, I can't resist.

"Establish bound[ary" is itself metaphorical and is surely not clearer than "define" itself. It is not as though undefined concepts are in danger of being infinite somehow (which is what your view, literally interpreted, seems to amount to). My complaint is not in the substance of what you say--it sounds right to me--but in how you state it (i.e., its clarity). You can probably state the gist of what you want to say just by stating: "A concept's definition describes all and only the items in the extension of the concept." --LMS

I find it interesting that Mr. 24 defines "i.e." and "e.g." in his explanation of why not to use definitions. :-) --LDC

Major problem
The define and describe rule makes up its own definition of definition. This is not fair. I have taken it out and linked to the definition page.Bensaccount 19:51, 5 Mar 2004 (UTC)

describe, don't define
Half-oppose, I guess. define is what you do to words. Wikipedia articles should be about things, because Wikipedia is not a dictionary. The only time you need to do something like definition is where there's some abiguity in the article title: but in this case you need only say "This article is about the city, for other meanings see Paris (disambiguation)", nothing more. Instead of a definition, always start with a solid description.

Example:


 * Anne Widdecombe is a Labour MP in the UK.

This is not a definition. There are many labour MPs. and Anne is just one. Further, when Anne ceases to be a labour MP, she will still be Anne Widdecombe. Instead, it is a brief description.

Harder example:


 * bad: homosexuality is sexual orientation to, or activity with, persons of the same sex.


 * better: homosexuals are sexually attracted to people of the same sex.

The definition is bad, because when we define, we typically have to spend ages hammering down a 101 loopholes: hence "sexual orientation to, or activity with". And someone will no doubt point out that the definition doesn't account for bisexuals, so to be accurate, we'd have to have:


 * even worse: homosexuality is exclusive (or near-exclusive) sexual orientation to, or activity with, persons of the same sex/gender.

The current introduction to homosexuality is:


 * hideous!: Homosexuality (rarely homophilia) is a sexual orientation or orientations characterized by romantic or sexual desire for, or sexual attraction toward, members of the same sex. The term usually implies an exclusive or predominant sexual orientation toward persons of the same sex, and is thus distinguished from bisexuality as well as heterosexuality.  In addition to referring to a sexual orientation, the term homosexuality is also used for sexual behavior between people of the same sex. However, it is worth distinguishing between homosexuality and same-sex sexual behavior.

If people make the same quibble about the better option: the description, we don't have the same problem. Yes, bisexuals are also sexually attracted to people of the same sex. At some later point, we can explain this. However, we don't need to put it in the introduction. Instead, we proceed on with the task of describing homosexuals.

Another point: with the definition, we had a weird article which dealt with homosexual orientation and homosexual activity: IE, both the sexual preference, and the sexual act. With the description, we have an article focused solely on the orientation: one thing. We can describe the sexual act at a more specific page title, such as homosexual sex. Martin 00:37, 30 Mar 2004 (UTC)


 * If you want thorough understanding you will have to base your associations on words, hence the definitions are necessary. Bensaccount 00:26, 30 Mar 2004 (UTC)

Definition: necessary and sufficient conditions
I come from a background in mathematics, where definitions are very fundamental. One thing I learned in my mathematical training is the value of making precise definitions. Relating to Wikipedia, the advantage of starting an article with a precise definition is that it makes it clear what the artcle is about (to both writer, editor, and readon), and also it forces the writer to develop a deeper understanding of the subject. Developing a consice definition for something can lead to a roadmap for further explanation.

To be blunt, if someone cannot formulate a proper definitition of a subject or concept, I think that person should avoid writing an article on that subject or concept. I know this "acid test" has stopped me from writing a couple of times, because in trying to formulate a concise definition, I realized that I didn't understand it.

In mathematics, definitions can be so precise that they become difficult to understand&mdash;textbooks often have to include a description or explanation of what a definition really means. I don't really expect Wikipedia to end up like that, but I think it does show that it can be useful to properly define something then spend the next few sentences elucidating the definition.

The second reason I bring up math is that there is a convention that a definition always implies necessary and sufficent conditions. As a simple example, consider the following, regarding a type of number hypothetically named a "graboid" number.


 * (1) A "graboid" number is a number greater than 0.
 * (2) A number is a "graboid" number if, and only if, it is greater than 0.

(1) is a description, (2) is a characterization. (1) only provides the necessary condition that if a number is a "graboid" number, then it is greater than 0. Whether the condition is sufficient is not specified. That is, it is not specified whether all numbers greater than zero are "graboid" numbers or if there are other characteristics of "graboid" numbers. (2) however, makes the meaning of a "graboid" number precise and can therefore serve as a definition. Notice though, that it is more cumbersome.

The convention in mathematics is to always assume the conditions in a definition are both necessary and sufficient; if the first description was in fact a definition, then it really means the second.

In the "real world" this "description = definition" convention is often assumed, but not always. In fact, depending on this ambiguity is an old trick of the advertising business. Does this sound familiar?


 * Your wireless phone contract provides you unlimited "weekend minutes." "Weekend minutes" are minutes you use on the weekends (Saturday and Sunday).

Of course later on in the fine print it mentions the further characteristic of "weekend minutes", that they only kick in after you have used all your regular minutes, even if you used them on the weekend. So the first weekend of your new plan, you happily talk all weekend, not realizing you're using up your prime minutes, and that you will be billed by the minute for every call you make on the coming Monday. The problem here is that a description of "weekend" minutes is confused with its definition.

I agree with Martin's sentiment that nailing down a precise definition sometimes just obscures the real meaning, but I also am inclined to believe that the opposite&mdash;being too vague&mdash;is more dangerous. In the homoexuality article, the better description is too vague, and frankly, it doesn't really convey information at all. The hideous version is not perfect, but it does make a clear definition, and also distinguishes it from other concepts that are often confused with it. In my experience, people understand new concepts better when they are presented with differences.

Also, there are many aspects to "homosexuality" that are inextricably intertwined with sexual orientation; it wouldn't be proper to write an article on same-sex sexual orientation without making that intention clear from the outset. Because "homosexuality" is complex term, with so many different connotations and meanings, an informative article really must lay out some basic definitions.

Anyway, my point is that if an article starts with a description, it should be a fair characterization of the article's subject rather than just a list of properties. Personally, I wouldn't want to find "a crow is a big black bird" as the first paragraph on any reference about crows; I'd like to know what specifically makes such a bird a "crow", and how they are different from other big black birds such as ravens.

Concepts vs. specific instances
Terms and concepts require definition, specific instances do not. Both however, require a characterization. The example of "Anne Widdecombe" above is a nice example. I don't know what it means to define a person, but I do know that there is probably something about Anne Widdecombe that sets her apart from other women in the UK, enough to have an encyclopedia article, and that salient characteristic belongs at the start of the article. A description of her physical characteristics might be a unique characterization, but the public at large is probably more interested in her political position.

Once again, this can be viewed as the "acid test" for writing an article about something specific, EG, a person, object, book, movie, whatever. I know if I can't come up with a consise expression of relevance, I won't attempt to write the article.

I hope these thoughts are of some use.Uranographer 06:33, 24 Apr 2004 (UTC)

Proposal to consolidate advice on writing better articles
At present there are many articles in the Wikipedia namespace that seek to give guidance on how to write better articles. I propose consolidating these into a much smaller number. On User:Jongarrettuk/Better writing guide I propose how these could be consolidated. The proposal is not to change advice, just to consolidate it. If I have inadvertently moved what you consider to be good advice that is currently in the Wikipedia namespace, please re-add it. I'm hope that the proposal to merge all these articles, in principle, will be welcomed. Of course, it may be preferred to have 2, 3 or 4 inter-connected articles than just one and would welcome advice on how this could be done. (In particular, perhaps all the guidance on layout should be spun off into one consolidated article on layout.) I'm also aware that putting lots of different bits of advice together may throw up anomalies or bits that people now disagree with (including bits that I myself disagree with:) ). I ask for support for the consolidation. Once the consolidation has happened, the advice can be changed in the normal way. Please feel free to improve on the current draft consolidation, but don't remove or add advice that is not currently on the Wikipedia namespace. If all goes well, I'll add a new Guide to writing better articles page on the 19th, though maybe some bits of the new article will need to be phased in over a longer period. I'll also take care to preserve all the archived discussion in one place. jguk 19:42, 11 Nov 2004 (UTC)

"Use clear, precise and accurate terms"

 * This is a hypothetical statement that I made up to illustrate a ploy commonly used in contentious articles. (I also posted it to talk:Patent nonsense but perhaps it would better go here.)
 * The writer wants to say "Beer is good for you" but he can't support the statement, so he says "Beer is nutritious". Well, beer has calories so it is nutritious in some sense. The statement really says:


 * "Beer is nutritious, depending on how you define nutritious" and the definition of "nutritious" is not clear from the context.


 * The ambiguity could be resolved by re-wording to something like "Beer contains nutrients" but the writer may insist on keeping his ambiguous but suggestive wording.


 * I can't find any explicit guideline or policy on this.


 * Could we say "Use clear, precise, accurate and unambiguous terms" and give some examples for the benefit of those who don't get it? 24.64.166.191 07:11, 17 Jun 2005 (UTC)