Has this view been repudiated? I remember when the original was posted, but since then, it seems that multiple active Wiki Gnomes have gone to the Please Please Do Categorize Every Page On Wiki side, in grand, sweeping gestures.
As someone who originally came down on the side of this page, I feel a sense of loss, because it seems that a few people have started categorizing every page on Wiki, without having resolved this debate.
The thing about wiki is that a minority can do as they please. There are people here who will neither be persuaded by rational argument, nor consider the views of others. This question has not been settled, and the category enthusiasts will categorize every page.
I'm all for wait-and-see. I figure people can categorize to their heart's content now, and if any categories really do decrease the utility of Wiki, we can go back and remove them later. It's easier to decategorize than categorize anyway, so if they really want to spend their time that way... -- Jonathan Tang
In theory, I agree. But the way Wiki seems to work lately is:
One person (or a few people) want to add something, and despite no consensus for it, they start adding it with great gusto
The majority don't like it and start reverting the additions
A massive Edit War ensues
The few start complaining they're getting picked on unfairly, invoking conspiracy theories and a "cabal" mentality on the part of the few folks in the mostly-silent majority that stand up.
A "mostly silent majority"? Then what makes you think the majority agrees with you? The other side also thinks of themselves as the few folks in the "mostly-silent majority". The truth is, both sides are vocal minorities. If the majority cared one way or the other, it would be obvious.
Valued contributors in the majority get fed up and leave
The result is just a big mess. After a few iterations of this, the majority just throw up their hands, and a bad (or at least marginal) idea carries the day.
The categorization question has been discussed to death already, and the result (judging from the several, mostly-redundant pages (like this one) seems to be that a few people think it's absolutely necessary, but the majority think it's neutral or slightly negative. It seems that some people (especially newcomers) don't have any respect for the Wiki consensus, and feel that because they can do something, they must.
This has something of the flavour of the Talent Pump.
My personal hope in starting this page is not to dissuade from categorizing of pages, but to dissuade from believing that every page needs a category. If ever you see a page that fits into a category, please do mark it. Likewise, if you ever think that there should be a category where there isn't one, go ahead and create it. But I think it's a bad idea to shoehorn pages into makeshift categories simply because they don't already have one.
The history of this page
This page was started in 2000, when Helmut Leitner added categories to about 200-300 pages in the course of one week. He also added about 20 new categories.
Helmut was reading every page in Wards Wiki (about 12000 at the time), and categorizing pages as he went. He added categories of a few kinds:
Categories describing people, places, or things: CategoryArtist, CategoryScientist.
Categories describing a topic or domain: CategoryPsychology, CategoryHistory.
Categories meant for use by Volunteer Housekeepers: CategoryNone, CategoryNoise.
Nobody ever doubted the good nature of Helmut's intentions, but many did take issue with what he was doing, and with the role of categories as an indexing scheme for Wiki.
In addition to the discussion, there were a number of votes taken. A Wiki Single Vote did not result in a significant majority favoring the continuation of category assignment, or stopping for more discussion.
Wiki Multiple Votes were also taken to see what categories voters thought were most useful. Voters were most likely to approve categories if they were specific (CategoryScientist, but not CategoryInsight). Voters also disapproved of housekeeping categories such as CategoryNoise or Category Misspelled.
It is good to keep in mind that completeness, correctness, and consistency are sometimes the friends of usefullness, and sometimes the enemies of usefulness. With categories, I think they are enemies. -- Stan Silver [inventor of categories] on Categories Discussion
I'm strongly agnostic about it. I doubt if anyone on Wiki knows today what will be useful in future, apart from a few categories already mentioned that definitely won't be. I would prefer that before too much work is lost statements began to be made about the relationship of categories in Category Category and above all that we're driven in our design by the evidence of real life successes. -- Richard Drake
Never add a category until you've experienced an actual need for it
Even then, you should carefully examine the existing Wiki Categories to find an appropriate match. The distinctions between categories should be clear and intuitive; otherwise categories in general will become less useful.
If you do find yourself tempted to add a category, ask yourself if it's a category that will be of interest to others. The more uninteresting categories Wiki has, the harder it will be to see the ones that really are interesting. (Certainly categories are unnecessary on deleted, empty, or misspelled pages, so don't use CategoryEmpty or CategoryMisspelled.)
Similarly, perhaps Wiki Gnomes should only categorize pages that are worth reading. Nobody needs a way to find useless information: It finds us by itself all too readily.
Category names should say more about the category than about the page it's on. For example, what's CategoryStrategy about? Who wants to see all the pages about "strategy"? What does mean? Categories add value by creating a collection of pages, so the name should make it clear how that collection is defined.
Refactoring Categories
Suggestion: Search for category. Pick a category and reverse search for it. Then decide for yourself if it is pulling its weight or if it should be refactored. -- Ben Aveling
If you find a category which has few backlinks, don't necessarily take that as meaning it should be removed. You might know that there are many pages which should have that category assigned but which, evidently, don't. What you can then do is go put that category tag on those pages.... although if in that process you discover that they already have a relevant category tag, but just different by name or spelling, then you have now discovered duplication to be refactored.
Must a category be so broad as to be practically useless? It seems that if there are at least a half dozen or so pages categorized under one category name, that sufficient reason for the category exists. It seems to me a category of from 6 to 20 pages is more useful than one which has several dozen to hundreds of categorized inclusions. When looking at a category list, one usually is focused in on the specifics included in the category name. Categories are not intrusive and take little space on a page. When utilized properly it is a far better way to navigate the wiki than Recent Changes and its relatives. It is important and wiki-like to add meaning to the wiki rather than being passive or subtractive to meaning.
Discussion:
I used to dislike categories intensely. I still do when they are added to what I consider very cool short pages such as Wiki Prayer, Rhetorical Question or Gk Chesterton On Wiki, not least for aesthetic reasons. I dislike the unstated implication that category bureaucracy and completeness is somehow more important than beauty and brevity. We need to cherish and protect the few examples of brevity we have on Wiki (he says, mounting his soapbox to produce yet another major counterexample). Note also that adding a category today to Gk Chesterton On Wiki would cause loss of this "signal": that the little deadpan comment at the end, made in December 1999, was by ... well, I'm probably not allowed to say, given that the person concerned hadn't implemented User Name for Wiki at that date.
Categories not Wiki-like
I agree. In my opinion, categorizing every page doesn't "feel" Wiki-like. It's not consistent with You Arent Gonna Need It, and it's not organic Piecemeal Growth. -- Falk Bruegmann
Hmmm... feel and wiki-like, not sure that works as an argument.
Categories add diversity
Having said all that, one thing that I have gradually begun to appreciate is that the category scheme, although started by one individual and currently being extended by a few dedicated categorical folk, is like everything on Wiki wonderfully multi-user and multi-faceted as it evolves. Topics have fallen by the wayside ... c'est la vie. Three categories on some pages, one on many, none on quite a few ... just enjoy the versatility. Categories that remain Prompting Statements ... why not? Thanks for those that try, thanks to those who don't, thanks to those that add, improve, remove. This surely is the Wiki Way.
Edit, Change, or Delete Categories - treat like part of text
Categories by definition should be considered the least authoritative text on a page, the stuff we think less hard before altering than any other. They're not authoritative because (normally) they've been put there by someone other than the real authors. I say this because they sometimes seem to have this solemn air of finality. They are there to be changed, commented on or joked about. As long as this is the attitude of the Wiki Community then I salute Helmut and others for taking so much time and trouble to continue and extend the experiment. -- Richard Drake
I dislike the unstated implication that category bureaucracy and completeness are somehow more important than beauty and brevity.
I dislike the implication that beauty and brevity should take precedence over improving access and use of the 12000+ wiki pages. By the way who (but you) said categories were bureaucratic? Finally what specific criteria would you suggest the community (systematically) use to assign or not to assign categories? -- Dave Steffe
Passer-bye: This discussion was so edifying until the word "you" started showing up in a seemingly snippy way. Then some passer-by had to go and point that out. An edit right in this area would turn a good page back around. I lost interest in it when it was turning personal here.
Don't worry, not all of the 12000+ are beautiful and brief. Maybe 120 are though. Those that aren't may benefit from additional asynchronous, incomplete and inconsistent indexing by multiple Wiki users. Well, strictly speaking the pages won't benefit. I'm willing to believe that future browsers of Wiki may find interesting pages this way. (Do we have specific testimonies on that? Even I'm looking forward to searching CategoryJoke.) The bureaucratic tries to describe my feeling so often when I see these things. It could just be my problem. I don't know how to answer the "systematically" question. I never have and would never expect to.
Wouldn't those that aren't benefit from the use of a systematic and consistent schema to categorize pages? What harm would come to the beautiful and brief pages? You have IMHO correctly identified the 'problem', namely asynchronous, incomplete and inconsistent indexing. It's a problem because some are unwilling. Why are you so against a cataloguing mechanism? -- DS
I don't see asynchronous, incomplete and inconsistent indexing as a problem but as a statement of fact. A fact of Wiki with perhaps as yet unrevealed but wonderful emergent properties. On with the experiment I say! Don't ask me to referee. -- Richard Drake
Good Alternatives to Badges and Categories
Rather than adding category and topic Wiki Badges to pages, I think editing text to transform phrases into Wiki Link Patterns is more useful, and leads to a richer Hyper Text than a few menu items at the top or bottom of the page. Many pages have been written about things that now have pages about them, and it is easy to add these kind of links by either compacting phrases into capitalized compound words, or by changing the wording slightly to create a Wiki link without changing the meaning. -- Nat Pryce
That doesn't really work for categorizing. How do you get an index of indexes? [er, CategoryCategory, anyone?] It comes down to this: if I want to find all the behavioral patterns on Wiki, someone would either have to (a) maintain a static listing page that would need updating each time a new pattern was added or (b) simply reference the CategoryBehavioralPatterns page. I think it's pretty clear which is a more elegant solution. The second is dynamic while the first creates an factorially increasing maintenance nightmare. Consider Pattern Forms - if I create a new pattern-style page I *also* need to update this page. With topics and or categories (who cares) all I need to do is reference CategoryPatternForm at the bottom of the page. Personally, I'm in favor of having as much on Wiki indexed as is possible - just as long as what is being indexed (i.e. referencing an indexing page) contains generally useful information. -- Robert Di Falco
PleasePleaseDontCreatePageNamesThatMakeStatements. A humble suggestion to consider page names such as Categorization Discussion or Categorization Disadvantages. This page's awkward name reminds me of the IRC practice of changing one's name to make statements when one has been blocked from discussion.
Arguments for the elegance of cleaner, shorter pages are much more credible when accompanied with cleaner page names. ;-)
Seconded!
(Edit Hint: perhaps this page and its sister Please Please Do Categorize Every Page On Wiki should be moved to Categories Discussion)
It would require the HintTaker to fix the scores of pages where it is linked by substituting with appropriate verbage or linkages in addition to refactoring the Categories Discussion page. The Don'ts link to 38 pages and the Do's link to 9 more, as well as the 3 other pages. That's a total of 50 pages. Any volunteers?
To get an impression about what is going on:
look into well established categories (CategoryHome Page, CategoryAuthor, CategoryBook, CategoryQuote, ...) and look for an overrepesentation of "A..."- entries.
look into formerly existing, but little used categories (CategoryCompany, CategoryPerson, CategoryJargon, CategoryExternal Link, ...) that now slowly develop.
look into new categories like CategorySoftwareTool, CategoryArtist, CategoryScientist, CategoryMagazine, CategoryWorld View, CategoryCommunication, CategoryPsychology, CategoryStory, CategoryJoke, ... that also develop slowly.
there is also the mixed group of software related categories I care for:
CategoryMethodology, CategorySoftwareDesign, CategoryTesting, CategoryDebugging, CategoryOptimization, CategoryMaintenance...
Generally I look at the Wiki empirically. I look at hundreds of pages and try to find a common denominator in the form of a category (if there isn't already one fitting).
Roadmaps, Searches as alternatives to Categories
Instead of making more categories, why not make more Road Maps? I find them more useful, and I bet newcomers do, too. -- Ralph Johnson
Yes, categories help users in their search for information. Their real importance is as a tool for those who build and maintain roadmaps. There is no sense in seeing categories and roadmaps as alternatives; they are parts of a system that have to go together. -- Helmut Leitner
After a while (perhaps after 10-15%, now: about 3-4%) of categorizing, I would write another roadmap, perhaps CategoryCategoryEmpiricalRoad Map to ease the access and to show the internal, multi-layered structure of the Wiki. Any number of such Road Maps could and should exist side by side.
I see my work only as a first step. It should at least
answer the question "what is in the Wiki empirically?" and
make the whole content accessible
I hope that others will follow and work more towards the content.
CategoryCategoryEmpiricalRoad MapConsideredHarmful, anyone?
Adding a level of order
First, the answer to the question, "What is in Wiki, empirically speaking?" is simply and only the page database (see Wiki List), technically speaking. So what are you really after?
Second, you shouldn't want the whole content accessible. You should really just want the actually interesting bits accessible.
Everything is accessible from Wiki List anyway. I just wanted to introduce one level of order and I didn't want to decide what's interesting. I just wanted to donate a few hundred hours of work to this Wiki, but this offer was turned down. I'm a bit disappointed. I will read the whole Wiki, but no one else will have any advantage from this. It's a waste, but not my problem. -- hl
Philosophical point: Wiki is not a governed entity. This vote thing is a good mechanism for collecting public opinion, but take it as advice, not as the rule of law. Stuff happens on Wiki because somebody tries it and other people talk about what they think of it. Sometimes people are persuaded to change their actions, sometimes not. On rare occasions people struggle with each other by changing a thing back and forth between (roughly) two states. We work things out with each other rather than giving and taking instructions (even though that's what it sounds like sometimes). Wiki runs on consent and consensus, not democracy.
By the way, I think that, on balance, you've done much more good than harm (if, indeed any real harm has been done at all). You took a large task onto yourself in order to benefit us all and I think that is to be commended. -- Phil Goodwin
Categories should help find stuff
The big CategoryJoke on this is: last week I asked for testimonies from people who had found specific categories useful for finding interesting pages. This I thought was the whole point of the exercise. What we got instead in Specific Testimony On Categories was testimony about how we've added categories over the years. Where is the testimony of the kind "I was crying out for some good advice on my new C++ project and when I searched Wiki's CategoryCpp I found something useful straight away"? It's never happened to me but I kind of assumed it must have happened to somebody.
We could perhaps build a rough ranking of existing categories, not according to a seven day vote on the predictions of current Recent Changes Junkies but based on years of real experience - the usefulness of categories to past searchers. Might this not help a little in designing and placing the categories of the future? On how much priority to give this kind of activity compared to other forms of refactoring? -- Richard Drake
Nowadays I mostly find stuff with title searches, which happens because I am looking for things I vaguely recall seeing here 2 or 3 years ago.
I would use categories more if they could be used to qualify full text searches. I'm unlikely to want every page in a given category. -- Dave Harris
It is also unlikely to want every page in a title search. Searches, Categories and Road Maps serve their own separate purposes.
Sorry that it took me so long to make this entry, folks, but I have to say the categories here helped me out of a jam a few years ago. I was trying to put together a set of white papers for a client. I remembered reading specific discussions on engineering process and process related matters such as requirements gathering and zero defect process, but I could not recall the specific pages, the authors, nor the words used to express some of these ideas. By searching through the categories I came across pretty much everything I needed to extract in order to bring my points into focus. Without the categories I would have spent a great deal of time hunting for stuff I needed but didn't know how to find. -- Marty Schrader
Categories should lead to enlightenment
With all these categories offered up in bulk, then withdrawn as suddenly, I was disappointed that my fave never surfaced, a CategoryZen. Although categorization is un-Zenlike, this could be a dandy category if assigned to all the pages in dire need of refactoring. By reading these pages (all of them) from beginning to end, the student of Wiki is bound to become enlightened. -- Walden Mathews
FYI - CategoryHomePage was the easiest thing to understand and explain when encountered. -- Michael Finney
Statistics might prove the effectiveness and utilization of categories
Lots of discussion here... I think this wiki would benefit from having a function that presents simple web server statistics on a monthly or weekly basis. We would learn which pages (categories or not) or searches are actually used, and could use this feedback for making, naming, and categorizing pages in ways that improve their usefulness. -- Lars Aronsson
Concur. It is possible that Ward or one of his lackeys already does this just as experimental data collection. In the mean time, this page is a Refactoring Candidate that I am somewhat afraid to mess with, being a source of "category noise" me own bad self. -- Marty Schrader
Pushing categories had at least these effects:
it made visible how many people had contributed to Wards Wiki (homepage count went from 300 to 2000).
it showed that some kind of overall refactoring of this wiki was still possible
it paved the way for the unification of the two categorizing systems (finally done by their inventor Stan Silver)
it practically established the use of CategoryHomePage (now showing 3300+ users)
it forced Ward to implement Recent Edits (because of 5000+ minor edits)
it forced Ward to implement page deletion (at the point of 200+ misspelled and categorized pages, aiming at a total of 500-800 garbage pages)
it forced Ward to implement "find pages of this category" with a nightly copy of the database, instead of scanning the "online" copy. (Category searches were preventing people from editing pages.)
So I think it was well worth the effort. -- Helmut Leitner
Look at Wiki Topics for the other categorizing system Helmut is referring to. -- francis
[I'd say that Recent Edits and page deletion happened entirely independently of categories. -- Jeff Grigg]
Would it be so horrible for some pages not have a category? Presumably, these pages are being linked from and to in a relevant manner. Pages without a category may simply be 'unripe'. In other words, the problem isn't a problem. Let us be patient, without Reinventing The Wheel. -- Todd Derscheid
The primary argument here isn't against categorizing pages, it's against creating arbitrary categories for pages that don't have an obvious category. The key word is "Every". Few would argue in favor of PleasePleaseDontCategorizeAnyPageOnWiki. It says that while categorizing is often good, some pages don't have a natural category and should be left alone. They can and will be found by links and searches.
I believe that:
everyone is in favour of some categorization,
some pages really should be categorized,
some pages really, really should not be categorized,
if you can't figure out what category a page should have, don't simply invent one - leave it alone.
It is not a contradiction for a person to agree with Please Please Dont Categorize Every Page On Wiki and to add categories to pages.
I am reminded of a quote by Michel Foucault from The Order Of Things.
Parallel: Fix Your Wiki
Categories only under this line if at all, unlike Page Names Category Category, Category Wiki History
See original on c2.com 