v1.0 1jul2001: First version
By Matt Webb, and Es Roudiani, starting at the Noodle King in Bethnal Green and continuing until the next day.
Is it possible to have a language in which it's really easy to explain concepts, but tough to say normal things? Like, the opposite to English where's it's tough to explain things -- look at the cumbersome nature of philosophical text.
Like: I can say why a giraffe has a long neck, it's because of evolution. But explaining what evolution is is really difficult.
Thinking about it in Dirk terms... Imagine inside people's heads it's all links and connections, all drawn out on paper. An idea is a pattern inside a head. Telling somebody an idea is a matter of explaining what the idea looks like and hoping the other person constructs it in the same way. In fact, lots of people have different patterns in their heads to describe the same idea.
(Recently, I've been calling this idea "engram squashing". A word is like a bullet, a compression of an engram that unfolds after passing the ears of the listener. A book is an argument, a nonlinear construction, a description of brushstrokes to draw a concept in your head [Vonnegut is very like this].)
A concept-easy language would let you draw a box round the pattern and trace it out, then hand the tracing to the other person.
The question is: is it possible to have a langauge like this, within the realms of the universal grammar switches? Although evolving a language like this might not have happenin, could we construct one? [Side question: What other possible valid languages are there which couldn't have evolved, but could now be constructed and used, and have advantages?]
A possible rule that would make a language more concept-easy is that before every paragraph, all terms must be described, and all those terms described too. (So a concept-easy language exists, in the sense that it could exist.)
But the rules have to be able to be applied to multiple paragraphs, across the semantic meaning of sentences. Can the rules from traditional languages act in this way?
Our general opinion: No, the rules can't act in ths way. So, it wouldn't be possible to construct a traditional language to be concept-easy.
So then, if we stick with English, is it possible to have metarules that act on the language to make it concept-easy? Even though they couldn't have evolved naturally, could one be created, so that untrained people could use it?
These rules would have to be contained in social memory. Some rules that are contained in social memory: the UK Parliament system, raising the intonation at the end of sentences.
General consensus: that social memory doesn't last long enough for this to work properly (the intonation is changing). So even if such memories could be considered, they'd never actually become part of the social memory, and as such only trained people would be able to speak or write it.
So, a concept-easy language could not exist with our current mental wiring. This implies that there's a limit on the complexity of concepts that can be exchanged by individuals.
Side issue: Imagine two Earths. The first is this one; the second is the same except the limit on the complexity of exchanged concepts is much higher. What would the differences be? Would Earth 2 be further advanced more quickly (or would they be fighting due to the deeper and more detailed insults that could be made?)?
Similar question: Imagine an Earth where the concept exchange complexity limit is lower. What would be different?
I'm only going to consider the effect on science. A question that needs to be answered first is whether the layers of abstraction in science are determined by the universe, or are legacy.
Definition of term: By science I mean the shared internal model with which we relate to the universe. [Dividing existence into internal and external, I tend to believe that interaction between the two is mediated by a model or map of the external that the brain continuously constructs.] Science is an abstracted model because it contains rules of thumb -- we don't calculate the wind from first principals, but choose to consider, say, pressure and temperature, which are both quite high level concepts.
If a scientific model was being constructed inside a single head, the abstraction layers can be wherever is most convenient. But to share a model, the reasons for abstracting have to be shared.
If the distance apart (distance being measured in complexity) of the layers is determined by the objective reality of the universe, then what happens if that distance is larger than the concept exchange limit? Does this impose a ceiling on the best model that can be constructed?
Or: Does it mean that individuals who are especially good will make the abstraction and be able to share the new layer itself, if not the reasons why? The new layer they share will have to be accepted into the shared model as gospel, only similarly minded individuals understanding for themselves why the leap has been made? [early quantum mechanics; Christ]
Or: are there several different ways of making the model, with different abstraction layers, and sometimes we choose one which is easier now but could be more difficult later, and a society with closer abstraction layers would end up building just as much of the model but rooting it on, say, music [thanks Olaf Stapledon] instead?
I would suggest that emergent properties fix abstraction layers in certain places (weather above gas; society above individuals; etc), but between these there are approximations which are flexible. But how this interacts with the complexity exchange limit I'm not sure.
All of which brings up another question: What is the feeling of having an idea which can be imagined perfectly clearly, but is too complicated to articulate?