The “techno-social gap” is literally a social gap that originates in technology. Closely related concepts include [[data_jail?]] and [[data_rules?]].

Usually dominant technologies or PublicDomain or OpenContent set [[precedent?]]s that become fixed. The most obvious example from physical technologies is railroad tracks, which only go to certain places and only accomodate exactly-compatible carriages. A techno-social gap is created everywhere the railroad doesn’t go. This is such a profound effect that it was used deliberately in war - shifting track guages so that it would become difficult for reconquered territory to be easily re integrated into the polity that was losing ground.

In wiki, MediaWiki is the dominant technology and [[anonymous_trolls?]] tend to use mediawiki format and consider it someone else’s problem to change links to ridiculous bad conventions like WikiWords.

Original Use of TechnoSocialGap

In 2007-04-22 Blog I (SamRose) wrote about a “TechnoSocialGap”. I did not see the existence of technological conventions, like CamelCase, as a deliberate tactic used by certain people to conquer and hold territory. I saw it as an emergence that happened in parallel with the emergence of other SocialMedia sharing tools, like blogs, and popular ContentOverCommunity wiki engines like MediaWiki.


I strongly disagree that CamelCase or WikiWords are bad, but I don’t feel particularly like debating the point. WysiwygIsntLinking. Another is that if you see text with spaces in it hyperlinked, there’s no telling what it’s hyperlinked to (perhaps a page, perhaps another website, whatever.) But if you see hyperlinked CamelCase – “Ah, that’s part of the LinkLanguage.” Regardless, I think this is about the extent of my enthusiasm in CamelCase’s defense at this point.


I agree that CamelCase is not bad. I think part of the difference of opinion lies in the fact that many people I come across are not using wiki to create LinkLanguage. So, they see CamelCase as “UgLy”, or getting in the way. These folks are often trying to co-author content under a familiar typesetting paradigm.

I think people are missing out by not trying to establish SharedMeaning? via LinkLanguage. I think that CamelCase was chosen in WardsWiki as a borrowed concept from SmallTalk programming language. I think that people who understand programming and work with programming languaged appreciate the LinkLanguage CamelCase concept. But, those who do not work with text files in a programming language paradigm seem to reject CamelCase. To people who operate mostly from a typesetting paradigm, all links are links. Appearance and readability from a typesetting paradigm are more important than the nature of links. The opposite often seems to be true for people who approach wiki from a coding paradigm. The nature of links is very important. Just as the nature of coding objects and what they link to, and being able to scan computer code and recognize all of those operators, and variables, etc is important for people from a coding paradigm.

So, I have come to see the TechnoSocialGap with wiki preferences as being rooted in the paradigm of how a person sees the content on the wiki page. CamelCase is programmatic reasoning. Many people involved in CommunityWiki are also programmers, and so are comfortable with programmatic logic embedded right into communication. The same is true for WardsWiki, and MeatballWiki, and MoinMoin.

Brilliant. :)

Will society adapt?

That is: Will society ever see code-points for ideas as worthwhile?

I personally think so, but it would require (at least) that the field of ideas be visible and understood before mainstream society can get into it. I am echo-locating for 3d engines that handle text as fundamentally as shapes (and not finding a whole lot.) Alice and wonderland is a starting point, in terms of people seeing the world of ideas. Can we attach code-points to it? IdeasLikeStarsAndSymphonies. Each star in the field has a number.

JaronLanier described words appearing in mid-air as transparent buckets full of the variability that the word attracts; the word “house” visibly conjuring 30 different houses of different shapes and sizes and ages, floating about within the container.

Words are containers, titles more specifically so. Our titles are intentionally short, because they are our language, so how do we make it clear? We CamelCase. WikiPediaIsNotTypical, nor is Wiktionary: every word is a title. Merely a LineOfThinking, to be sure, but I think SamRose is onto something: Typographics vs. LinkLanguage, and WysiwygIsntLinking. Room for something different; Will society then understand? Or can they not take in LinkLanguage?

LinkLanguage may be totally different should the web revolutionize to a new platform, like a DevelopersVirtualWorld. Some kind of CyberSpace dramatically different. 30 years, 40 years, but less if people are enterprising and things work out.

TechnoSocialGap. Someone invents cars, and the people on horses are confused.

I need to ask the author, if he is indeed RainerWasserfuhr, (perhaps an email to Rainer is in order,) and if he is speaking of “anonymous trolls” in the pejorative or the honorific..!

PlainLink can work as a LinkLanguage If page titles are chosen carefully, and also avoids most of the troubles discussed on both sides of this argument.

if you see text with spaces in it hyperlinked, there’s no telling what it’s hyperlinked to

Strict PlainLink would not allow external links to be hidden behind words. An external link would be the literal text of the URL in question such as or and could also be highlighted with a different color through the stylesheet.

The environment can also be coincidentally WYSIWYG even using just regular text input (though the input box must not be allowed to arbitrarily insert carriage returns to artificially break lines) when used in combination with PlainTextWiki while still allowing the community to build a natural LinkLanguage.

For instance, the phrase anonymous troll is already understood to be a title, so why not just link it even though it contains a ‘SPACE’ character?


Your ideas about programming languages have some merit, but I wonder if they are less solidified than you imagine.

For instance, I remember reading about an Emacs implementation (I thought it was written in Common Lisp, but can’t find the reference right now) that allowed spaces in the command names instead of the usual ’-’ character; and with the commands named (titled) to make them appear to be english phrases.

On a similar (in my mind) note, are the command parsers for interactive fiction (text adventure games) that deal with verb, noun, adjectives, adverbs, prepositions, etc. without requiring spaces be removed…

Anyway, I’m only trying to help us grow. Maybe my ideas are incorrect. Let me know the specifics if you can see what won’t work.

Thanks, – PatrickAnderson

I think there is something important to be found in the notion of separating “content” and “presentation”, in having “source code” and “application”, “script” and “movie”, “raw page text” and “rendered html”. The idea is something natural to programmers, and it’s very comfortable to have access to the “real thing”, the data that is actually stored in the wiki, sans all the transformations that make it look like a wiki page. That’s not to say that “presentation” is not content itself, not valuable. A movie script is not a movie. Yet it’s the “heart” of it.

I find it very hard to explain to “non technical” people. There is no problem explaining it to “technical minded” ones, even if they don’t know anything about computers or information architecture – it seems that being able to look at blueprints and see the machine in them, or to look at a music score and hear the music is a related skill. And no, not even all mathematicians have it, so it’s different from symbolic thinking.

Lion: I think that society will see code-points for ideas as worthwhile. But, I think that many people basically demand that the concept be reframed and reformed into something that is very simple to use and understand, yet still retains the same power. For instance, the interface of vs. the UNIX command line version Wikipedia:Pipeline_(Unix) not an exact one-to-one comparison, for sure. But, it demonstrates the types of interface and functionality that resonates with most people.

It’s also a kind of human struggle where some people want to work with what works really well right now, vs. people who want to focus on what the utlimate solution could be (but that would take huge amounts of of work and energy to create, usually).

Patrick: I agree with you that it’s not “solidified”, for sure. My description was a simplification. But my overall point comes from looking at who I have met that is advocating CamelCase, and who is not, and my best guesses as to why.

I think that in particular, in wiki, starting with WardsWiki, and the developments that happened within the WikiEngine that powered WardsWiki, were seen at the time as radical programmatic evolutions to static HTML webpages. I think they (WardCunningham and others) modeled some of their functionality and concepts after the ObjectOriented nature of the SmallTalk programming language.

Yet, as I mention above with the example of YahooPipes, for instance, I think that a significant amount of people basically demand that the functionality of something like LinkLanguage be refined into something they can understand and use in under, say, five minutes.

I agree with Radomir that there is a certain type of thinking inclinement that will understand blueprints, music scores, and computer code.

I don’t think it’s totally symbolic thinking either, although I think that is part of it. Instead, I think it could be rooted in a really good ability to recognize patterns, and even more importantly, the inherent ability to imagine those patterns played out in some form of reality. Also, everything mentioned in this type of perception seems to be tied to “systems” and scales within systems.

Patrick,see also notes at from RadomirDopieralski, there are definitely unresolved issues with all PlainTextWiki solutions so far.


The first issue is:

Impossible to link to pages with link text being a pluralization or other transformation (tense, mode, gender, whatever) of the page title

It is not Impossible to link pluralizations or transformations in PlainLink, it is just not automatic.

It is not Impossible to do this with CamelCase either, and it is also not automatic.

It is a point that needs to be addressed generally, but is not specific to PlainLink.

CamelCase is actually worse in this regard because it is case sensitive.

For instance, AnonymousTroll?, anonymoustroll, anonymousTroll, Anonymoustroll, ANONYMOUSTROLL are all different (and many permutations I didn’t list).

Obviously AnonymousTroll? and AnonymousTrolls? are also different in CamelCase.

Furthermore, PlainLink could link many variations such as anonymous trolls, anonymous trolling, anonymous troller, anonymous trollobite, etc. by using a regex pattern that looked for the text “anonymous troll”, but allowed trailing alpha-characters up until punctuation or whitespace.

By the way, GaGaParser does not implement PlainLink, it sound like it does something like “make every word on the page a link even if a corresponding page title doesn’t exist”.

PlainLink is different. Text only becomes a link in a PlainLink when that page title already exists. This means when new pages are created, the text of the entire database must be searched for occurrences of the new page title, and the pages containing that text must be re-rendered.

The second issue is:

But not when you have to run together several words. Single-word links aren’t usually very informative, and also are against w3c recommendations.

GaGa requires ’_’ between words to make them link. PlainLink links any title in the database no matter the punctuation or whitespace within the title.

If the administrator of a PlainLink wiki doesn’t want single-word links, he can just require users only create page titles of two or more words.

two points in addition to everything here. 1) I describe the style of linking in wiki to the inexperienced as unique identifies for an idea. They seem to get it in many cases. I just ask them to go to google and type in the two words they want for an idea and hit enter - then asked them to CamelCase them and see what the get. Many folks get it at that point. Wiki itself is a TechnoSocialGap for many folks. Younger folks are more naturally getting it and I think CamelCase LinkLanguage will be a part of that also. My two cents :-)


I quite like your suggestion about how to introduce CamelCase and linking via google. I think some of the Audiences I deal with will find this helpful.

Your idea that a CamelCase name is just a unique identified for a page is absolutely consistent with what Ward has both stated and written many times in the past. I’ve also just published and presented a paper in which I argue that a pageName can be thought of simply as an index into the wiki’s namespace. Most Audiences immediately understand this when I link the Wiki page retrieval algorithms to the use of an Excel spreadsheet reference (such as “B3” for the 2nd (“B"th) column, 3rd row) that happens to contain a block of text like a page. I’ve even seen heads nodding when this is extended to think of the InterMap reference as a type of Work Sheet reference, thereby converting the usual spreadsheet matrix into a three dimensional array. This approach is even a great way to lead much more sophisticated Audience to more technical extensions of this model. For example…

  • Mathematical audiences who are comfortable with EnDimensionalRaggedArray? constructs still appreciate the initial introduction of the spreadsheet metaphor, as a warm-up prior to introducing the idea of mixing different numeric and alphabetic bases of the usual polynomial expansion in order to increase the information density (which dramatically reduces access and sort times for those of us that work with extremely large data structures).
  • Technical audiences who are familiar with SQL and other generally case-insensitive methods, begin to understand that case-sensitivity vastly increases the available address space to the extent that it becomes very profitable to avoid limiting oneself to what “people” (the unpredictable, unreliable carbon-based processors that they are) consider “pretty”. This generally makes it easier to increase the work assigned to the competing silicon-based processors, and optimize “profit”.

:-) There have been complaints lately that CW is too quiet. I certainly feel I’ve been too quite, but that’s because I generally avoid disagreeing openly in order to save time for more enjoyable pursuits. Maybe my profit allusion and stated preference for silicon based logic will help provoke a bit of noise for our Lion. :-)

Patrick, I noticed on PlainLink that you had talked about trying to build this by extending an existing wiki engine. I wondered if you had made any progress on this? (Will post the rest of my question on PlainLink)

I did not create this page (It was hlfx*, /me usually: p*, but I liked the railroad metaphor.

I think the author was PatrickAnderson, unless I am mistaken. This was a good thread and page. It “woke up” CommunityWiki :)

No, that wasn’t me. I don’t even understand what some of this means.

For instance:

tend to use mediawiki format and consider it someone else’s problem to change links to ridiculous bad conventions like WikiWords.

It almost sounds like [s]he is saying trolls use double []‘s for linking, and those same trolls assume someone else will change them to WikiWords (as though WikiWords should have been used in the first place?), even though this author clearly dislikes CamelCase. Confusing.

I might be mistaken, but it seems that the original poster is talking about a gap between two incompatible technologies, and how creating of such gaps artificially is often used to an advantage in a political struggle. Often you can create such a gap by adapting and slightly changing the standard – then you create a split between the two “standards”, and if one of the divided parts is large enough in comparison to the other, the other can slowly wither and die. This is especially useful when you have some control over “the majority of the market”, like Microsoft has – actually, this technique, called “embrace and extend” has became the characteristic of that company.

  • Radomir… Thanks for taking the time to add this explanation. It seems obvious now that you’ve stated it so succinctly and its even consistent with what I’ve observed, but I just never thought about it before. – Hans

Wait, I think I know who this is:

He has been messaging me on Jabber, asking me about some of my ideas related to stuff on wikis. And, he’s from Canada. And he has info on his wiki page about being a self-proclaimed “troll” (?) Not that it really matters who it is from.

Apparently, we have been Trolled! :D


Define external redirect: data rules precedent EnDimensionalRaggedArray anonymous trolls AnonymousTroll data jail AnonymousTrolls SharedMeaning

EditNearLinks: MediaWiki WikiPediaIsNotTypical MeatballWiki CyberSpace GaGaParser YahooPipes UgLy WikiEngine WikiWords CamelCase GaGa WardCunningham PatrickAnderson ObjectOriented MoinMoin SmallTalk WardsWiki