Wikipedia: WikiProject Concepts/Recap

Larry: I've tried to summarize your concerns about Wikiproject Concepts (wpc) and provide answers to your questions. Tell me if I have misunderstood.

Seb, I'll just intersperse some answers--can't pretend the answers will be perfect, as I'm trying to budget my time here. --LMS

You ask of what use would a conceptual hierarchy be if Wikipedia articles had an explicitly delineated logical structure of links such as is proposed in wpc. I think that such a hierarchy would be completely redundant with respect to the encyclopedia itself. As such, the hierarchy is only useful inasmuch as Wikipedia articles covering theoretical constructs lack an explicitly delineated logical structure of links.

So, do you want to say that Wikipedia would benefit from this project only to the extent to which it "explicitly delineates a logical structure of links"? If so, why not just add that structure to articles? Why develop the structure independently of articles? That's one point that really confuses me.

You seem to say that building a conceptual hierarchy will result in inaccurate simplification. I agree if we're talking about describing the real world. However, I disagree if we're talking about theory. Correct me if I'm wrong, but I believe that a theory is by definition a simplification that is intended to make it easier to think about something. In that respect, a theory should be for the most part describable with hierarchies of concepts.

Most encyclopedia articles do not concern mere theory (e.g., mathematical constructs). So does WikiProject Concepts concern only what you call "theory"?

You contend that it's a mistake to try to shoehorn content into a hierarchy. I clearly understand why, if you're referring to a treelike hierarchy (i.e. a single parent per concept; then a single generalization is accessible from a concept - bad) or to a hierarchy that seeks to classify real entities (edit wars galore). However, in the case of a hierarchy of ideas that allows for multiple parenting, I fail to see what the mistake would be. Perhaps you could give a concrete example of an idea that one could not situate within a multiple inheritance hierarchy without falling prey to Murphy's Law?

Well, then I have much less of an objection along those lines. But then the project looks a lot more like Wikipedia itself.

You ask how something like wpc could give clear, helpful access into the content of Wikipedia. Here is one example. Suppose I am a budding mathematician and I have come to suspect that there is some kind of commonality between the cube and the regular tetrahedron. I don't know exactly what the name of that curious property is. Then I could look up the generalizations of either concept and quickly find that they are both symmetrical geometrical objects. From there, craving knowledge about symmetry, I could consult the relevant Wikipedia article. Of course I could have found this information and/or link in a Wikipedia article describing either concept, but this likely would have taken more time. See cube and tetrahedron, which are rather short articles by encyclopedic standards, to see what I mean. Looking up longer articles would arguably take more time.

In other words (I guess), you want people to have short, clear articles about difficult concepts. I totally agree. In fact, although our mathematics section is one of the best so far (because some actual mathematicians have been at work on it!), one aspect of it that I've deplored has been the fact that concepts are often not introduced as simply and clearly as they should be. This does not necessitate that we provide explanations in separate articles, though--why would it? That's what encyclopedia articles are for, after all!

You ask if it would not be better for the contributors of articles to think for themselves about the link structure between what they describe and other stuff. I agree wholeheartedly. A contributor could both write an article on a topic he knows well, and lay out its conceptual neighborhood structure to encourage inflow of interested people towards his interests.

What I fail to understand is why anyone would do that using the WikiProject Concepts schema rather than just good old Wikipedia articles themselves.

You mention a threat of internecine warfare about what the "correct conceptual scheme" is. This is sure to happen if we seriously try to classify real-world entities. However, in the world of ideas, I don't see how this can happen, as the proposed methodology can accomodate any number of (possibly competing) theoretical schemes. Is there room for disagreement about mathematics-like definitions? The usual wisdom there is, if you don't like my definition, you make up your own. Of course there will be namespace competition; a simple numbering scheme could do the trick of disambiguating names. --Seb

OK, so you are limiting "concepts" only to "theoretical schemes," whatever those are exactly.

I'm thinking more and more that you're mainly interested in developing a web of concepts, which doesn't have any clear use for Wikipedia. Again, I share your interest in that sort of project. Moreover, I'm not entirely averse to your starting to develop it here on Wikipedia (see [1]). But, short of some really good arguments that Wikipedia would clearly benefit from this project, I think it would be better to find another wiki for it. There are many wiki farms out there, and if you're a hacker you can easily set a wiki up yourself. If you ask Jimbo Wales nicely :-), he will probably set one up for you on http://www.wikicities.com/ (which right now for some bizarre reason just redirects to Nupedia). --LMS