User:Aya/Wikibooks/An analysis of wiki-based communication

From Wikibooks, open books for an open world
Jump to navigation Jump to search

An analysis of wiki-based communication

This is just an attempt to shift the more generalized areas of A critique of Wikibooks into a separate document. The word 'wiki' in the title should be considered synonymous with 'MediaWiki'. I may include other wiki systems over time.


MediaWiki is a recent and popular implementation of the 'mind-map' concept. I've seen this as far back at the late 1980s, designed for use either as a personal database, or a means of explaining language. The success of the internet means this can now be done collaboratively with people from all over the world, but this introduces new problems.

With a single user editing the whole thing, they can use whatever terminology they desire, but when working with such systems collaboratively, shared language must be agreed upon. This is, in effect, the system-wide NPOV policy, which is arguably the only policy you really need. The effect of applying this policy to the definitions of words and concepts you might find in Wikipedia, you end up having no choice but to agreed to shared language.


What are the problems with the use of language?
  • Obvious things like the existence of many incompatible languages and dialects thereof.

These have mostly been addressed by splitting wikis into language codes, but even within a single language code, many dialects exist. q.v. dispute on UK/US English, amongst various other arguments on m:Requests for new languages. When is a dialect considered unique enough to warrant its own language code?

  • Other problems with both written and spoken language

e.g. category logic errors, metaphor vs practical definition, meaningful statements.

Note that the use-mention distinction has already been pointed out somewhere in w:WP:MOS. In fact, this may already contain many solutions.

  • Intonation in spoken language

The practise of changing one's tone of voice when saying something, to hint at an alternate interpretation of what they are saying, doesn't translate well when written. I notice a lot of subtitled TV shows use the tokens '(!)' and '(?)' at the end of a sentence to imply a 'sarcastic' tone.

  • Body language

i.e. actually doing something rather than talking about something

The ultimate language, which if taken 'literally' lacks any ambiguity, although this is probably beyond the scope of a wiki.

Could a natural language ever evolve into the use of fully discrete concepts?

By which I mean, all ambiguity is removed.

Modern high-level computer programming languages are not far off, since they effectively let you define your own language of 'identifiers' to keep it meaningful. Low-level languages that use numbers to represent operations and data (verbs and nouns) might be more accurate.

I don't think this type conversation will ever become commonplace:

Fred: 287 10 1879?
Bob: 3!

That is to imply that each number represents a fully discrete concept, although many geek-conversations do approach this level of disambiguation.

What are the conceptual differences between nouns, verbs, adjectives, etc?

It seems you can translate between them quite easily. Might be worth investigating w:Lojban to see what they came up with.

Are there examples of natural languages which do not have these distinctions?

I'm wondering about pictographic languages like the Kanji. Need to check this out.

Also, languages of other life-forms. Body language?

What exactly is the canonical mapping of English words to their referent?

e.g. OED, Brittanica, etc.

I use the term 'dictionary' to include 'encyclopedia'.

But what happens when competing dictionaries disagree, or provide alternate meanings that others do not? How can a paper dictionary ever be future-proof? Is an evolving dictionary like Wikipedia the only solution?

Are all the Mediawiki wikis actually trying to achieve the same thing in a different way?

e.g. scope overlap with Wiktionary/Wikipedia/Wikibooks

What words are dangerous?

An argument from w:Talk:Words to avoid even suggests that the verb 'to be' is dangerous, since it argues that other verbs are almost always more accurate. This may be true, but it's such a common verb, it's not practical to simply stop using it. Perhaps its meaning just needs to be tightened up. e.g.

  • X am/are/is Y

If 'X' and 'Y' represent concepts invented by people, which are generally well-defined to be objectively distinguisable regarding their accuracy. q.v. boolean algebra, set theory, etc.

If 'X' and 'Y' are, instead, commonly used words to represent philosophically undefinable concepts (e.g. truth, morality, etc.)

There ought to be some sort of measure of meaningfulness, based on how many other concepts are required to understand in order to understand it. q.v. concrete nouns.

  • Is X Y?

Again, depends on X and Y, but does this really mean:

  • Please define X in terms of Y, or Y in terms of X.
Should self-contradictory words be deprecated?

For example: the word 'atom', literally means 'indivisible'. Popularly, it means one of those things physicists keep banging on about, but those are not indivisible (metaphorically speaking, I guess). Therefore, just by using the word 'atom' in the latter sense, are you implicitly (and erroneously) saying that an atom is atomic? The flip side of this is you can quite correctly say "an atom is not atomic". Does it strike anyone that this might be somewhat confusing from a typographical perspective? Do we have some subconscious desire to deliberately obfuscate our language?

Completely unfounded assertions[edit]

For want of a better place for now.

The Wikimedia projects should define the MediaWiki software, not vice versa.

This is probably already true, but having investigated the latest betas of Mediawiki, I feel a case of w:feeping creaturism coming on shortly. I guess the project can cope, since it runs more off wetware than software.

Need better distinction between words and concepts.

Wikipedia has already bumped into this problem, and created disambiguation pages to solve it. I suspect in many cases of words and word-phrases, a disambiguation page is not necessary, but as the amount of discrete concepts increase on Wikipedia, so will its ambiguity.

Why couldn't a wiki contain a master index of concepts, then attach the various typographic and phonetic language 'codes' (i.e. written and spoken words) commonly used to refer to each concept. Problem is that in order to define the concepts, you need some language in which to do so. I guess real-life photos, sound-samples and other sensory cues could be substituted for text where possible, since they arguably provide more meaning than text in any case.

I argue that this is the only goal of language. It would seem that almost everyone agrees on the concepts of the universe, but merely argue about the language used to refer to them, and the boundaries of the conceptual cases that they cover. For example, the concept generally referred to by the word "God" could perhaps also be referred to by the words "the laws of physics", in that they both refer to a conceptual entity which is allegedly responsible for the creation of the universe. Does this mean that the two phrases may be used interchangably? Perhaps not, since the "God" concept implies something of a vaguely humanoid form, but if I apply this to the first sentence of the Old Testament (UK edition), I end up with:

  • In the beginning when the laws of physics created the universe, the earth was formless and desolate.

A physicist might argue that the earth did not exist at the beginnings of the universe, but I suppose the word "when" in that sentence could be reasonably argued not to read as "at the exact instant that", but rather "during the period when". Point is, it's arguably correct.

What if a physics student studying gravity was asked by student of religion:

  • What did you learn about today?

Allowing for a certain amount of metaphor, could he, in all honesty, reply:

  • And the Lord spake, "what goes up must come down".

Is this merely a translation of the same concepts into a different dialect of the English language?