“LinkLanguage” is the language formed by familiar links.
Have you ever seen a page change drastically? One that people linked to a lot?
Have you ever wanted to link to something, but been a little nervous about doing so, because it might change or already have changed? The difference between having to look at current versions and perhaps choose to link only to a specific version, or just trusting that the current version supports the point that you’re trying to make, is very significant. Count the mouse clicks and page reloads to just link to some known and memorable page name versus having to do this research and link to something more specific - it’s astonishing how much more difficult it is to link without some trust or faith in the process by which the page is edited, maintained and moderated.
“TrustedLinkLanguage” is link language that people trust linking to, because they know it won’t change on them (“ContentSwizzling”) so significantly it won’t serve the communicative purpose it used to serve. Such trust could be misplaced, it could be totally unreasonable (expecting a long story about some personal event to remain on a technical reference wiki page for instance would be unreasonable), it could lead to a lot of counter-productive outcomes and reduce trust for a lot of other people (if you expect a page on evolution to debunk it and explain only scientific creationism, say).
For example here’s a story that might disappear and shouldn’t be expected to remain since it’s in FirstPerson and accordingly useless as a community wiki page, especially without quotes or attribution:
:I reworked a page on Ward’s Wiki, Wiki:OffTopic, which I was proud of. I liked what it said, and I link to it a lot, because it means particular things. In particular, it includes notes on how some OffTopic is actually helpful and necessary, because people need to be able to bond, in order to work together. I felt that the argument resulted in good understanding. However, Ward’s Wiki is a chaotic place. Someone may double-delete the page (link to the better name for this?), rendering my links useless. I hesitate to link Wiki:OffTopic.
InterWiki communities often compound the problem- the person who is linking may have some distance from the community that is maintaining the linked wiki. That means that the person is less likely to trust that the LinkLanguage is valid at least for their purposes. The less disciplined they are about their use of terms, the more likely they will mean something different for someone else.
A TrustedLinkLanguage usually results from a mature community around some topic that builds an accompanying ManagedWiki, or perhaps even static pages that can be challenged by some documented process. This can be considered a sort of “Semantic Infrastructure” effort for ThePublicWeb, and many such efforts already exist now - like OpenCyc for instance or Wikipedia itself.
Wikipedia is the most obvious and interesting case. People often slant articles to their point of view, then point to them from other materials. Adversaries may then slant them towards the opposing view. Ultimately others will become involved in the TrollWar? that invariably results. However, when the BattlefieldOfIdeas? settles, the most stable and worthwhile points usually remain. So when making non-controversial claims you are usually best off linking to Wikipedia, but when trying to convince people of ancient astronaut theory you may be better off linking to UFOpedia instead. People often link to specific versions of Wikipedia pages, especially on talk pages of Wikipedia itself. It’s sometimes suggested that only some “approved version” should be displayed to the public until a replacement/update can also be “approved” but this would remove the motivation to update pages and empower the clique who did “approvals”.
It isn’t necessary to have a real-time devoted fanatic army of many factions of trolls trying to score points on each other by removing inaccuracy or bias (or if you prefer inserting it!), as Wikipedia has, simply to achieve pages improvable by their users.
Even static pages are modifiable in theory - if not then we are stuck in one particular era, so let’s call these a NearlyStaticPage. An automated way of accessing the pages, guarded by a community of people with some common discipline or criteria, perhaps an InstrumentedWiki? or some similar mechanism, could be easier to manage in some circumstances than a typical wiki.
Technology to do this may involve WikiFeatures:TemperedPages? (see WikiFeatures:IdeasToPlace #118.) Static pages would include links to freely editable versions of pages, likely in a ScratchWiki mirror, or in a CommunalWiki. You could probably also automatically copy the page to another wiki, or download it to your hard drive. For instance it’s very easy to install the HaloExtensions? of SemanticMediawiki? with a single installer, and run a local wiki with excellent features and performance. Many of these could then cooperate, people sharing with many others on a bilateral basis and determining which of many versions was authoritative by algorithm.
This effort to build ManagedWiki or static pages would be part of the PublicRefineryProcess. A NearlyStaticPage to come out of a process would be well respected, and frequently linked and visited. Refining pages, and maintaining TrustedLinkLanguage would likely be a SuperordinateGoal of a group, which might consider itself a subset of an online community, or might have its own separate goals.
When building new extensions to FOAF, for example, we see a similar thing happening: They want to make an extension to describe pets. Now, how do we know your dog is a dog? They will use a tag to describe the species of the dog, and the rag will be linked to WordNet. WordNet provide us with a “link language” for ordinary things – it turns our usual vocabulary into a static, slow-changing, linkable vocabulary of NearlyStaticPage choices. Ones that closely align with expectations.
I’ve moved this page out of LinkLanguage. It needs serious reworking.
Yes, I’d like to have us avoid some of the fallacies about language, such as the WordNet idea described above. This will take some work to both describe and avoid, especially since this kind of thing is very common…
It is becomming apparent that there is a great deal more to TrustedLinkLanguage than just knowing and appreciating the concepts documented in the linked material. As more “web services” applications emerge, the approach of passing arguments to the processor that allow it to perform a specialized function and return a TrustedResult? will become more prevalent. This is likely to cause a real ParadigmShift in the way we think about computation and communication.
This is even more problematic in the case of Wikipedia - it’s true that you can always view the history, but if you make a point and you want to support it with Wikipedia and your adversary just see that and goes there and subtly changes the page then it can create so much chaos. I see it as inevitable that at some points we shall start to link to specific versions of the pages and only from there let people to go to the current page.
Reworking somewhat and updating. Dealing with the Wikipedia case, linking to versions, OpenCyc as a better example than WordNet, etc. The main point seems to be that things become “Trusted” only because of the people devoted to a particular predictable editorial process that they follow in common and in which they accept appeals, challenges, etc., and would update a page reliably if this process resulted in a successful appeal/challenge/version. So the word “Trusted” is probably too general, as is “language”, and maybe we are simply talking about a VeryPredictableLinkSchema - not a whole “language”, and predictability not “trust” as the goal. You can predict for instance that Conservapedia will be anti-evolution and Wikipedia pro-evolution, but that doesn’t mean you necessarily “trust” either one.