(Up to: Comms As An OOBE Laws Vs Communication )

Is it possible to construct networks that have no set protocol? Only, perhaps, a way of describing that protocol, like mark-up.

This assumes that protocols have two tasks:

1. Define WHAT data can be passed around
2. Define HOW that data can be passed around

Point 1 may also encapsulate data constraints, such as typing and length limits.

If we remove the need for point 1, can protocols for data specification (rather than language specification) "evolve" or emerge, through use?

If we remove the need for point 2, can language protocols emerge through "experimental" clients? i.e. can they develop a language between them. This was tried a few years back, by Sony I think, to let software clients describe "locations" between themselves, without setting any rules. I'll try to dig out a link.

The thing is that humans communicating through a third party, such as a computer, are not too bothered about point 2. But point 1 does have an effect, as it defines and shapes the data that they are manipulating. Furthermore, why should the underlying technology care about what data is passed around, if it is relevant to the users rather than the machine?

I think that getting away from typed data will also help in this regard - by separating machine protocols from human inference, we can make more flexible, natural networks.

Other things that may or may not work on a similar basis. Of course, this could be considered merely a layers-of-abstraction problem, so the idea can apply to many, many things. :)

• Money - minted currency is a form of standardised exchange platform, but there are many, many layers of economy above that - interest rates, inflation and exchange rates all server to "adjust" the actual "value" of the money, and therefore the value of the otherwise constant goods and services around you. (A spanner is still a spanner and does spannering, even if it costs twice as much. The only thing that changes is the economic ground beneath it.)

• Human Language - we have vocal chords, we can speak and we can hear. But language itself is not fixed. The languages we have today are an evolved amalgamation, formed from thousands of years of experimentation, crystallisation and slang-twisting. But we tend towards common languages that are quite, quite abstracted away from any original reason for sounds becoming words.

See also the ThoughtStorms page of the same name.

(See also: Humans Don TCast Information Mapping Emergent Conversationalists )