[Atlantium]

IQ testing is not how anyone measures intelligence and hasn't been for quite some time. I suspect the problem lies here: Intelligence requires a multi-part definition, to use your analogy, much like swimming. Swimming is defined as the action of moving oneself through water using your limbs. An intelligence definition should include the ability to extrapolate, comprehend and innovate information accurately as far as we can understand the world. The more precisely one can do this and the more depth of knowledge one has the more intelligent one is. A tricky thing to define.

[derefr]

> An intelligence definition should include the ability to extrapolate, comprehend and innovate information accurately as far as we can understand the world.

What? "Intelligence" is not a psychology jargon term; it's just a word, defined by the way lay-people put the term to use in conversation. Language is used to communicate; we define words (jargon excluded) as what the majority of people understand them to mean.

When jargon and regular words collide, the lay-definition wins, and the jargon definition gets lost. (See e.g. "begging the question", which has become a lay-term for "something that has an obvious corollary or unstated flaw" rather than its meaning as a jargon term in logic.)

That's why academics invent terms like "g" — to make sure that the jargon term has no lay-term it's colliding with.

[Atlantium]

Lay-people have incomplete information, therefore even if they "win" with common usage their opinion can be casually disregarded. If you are suggesting that I give in to the ignorance of the average human simply because they have numbers you and me have a serious irreconcilable problem.

[derefr]

I am suggesting that the purpose/point/craft/art of communication has the goal of packing the model in your head into a portable "universal standard" form, so that it can be unpacked as losslessly as possible, in minds with models as different from your own as possible.

The concept of a lingua franca is that there is an optimal encoding for information conveyance, if you're recording that knowledge without knowing the target audience in advance; and that that encoding is the one that uses the language, the dialect, the terms, the idioms, the mode of speech, etc. that — at least measured by current world trends — will confound the understanding of the fewest such potential audiences.

Or, to put that another way: if the people you want to communicate already think like you... do you really need to say anything at all? They can probably independently come up with the same thought you want to convey. Communication exists to bridge gaps in understanding, to create mappings between non-isomorphic mental models where those mappings aren't already implicitly embedded in the models' shared structure. Communication is work done upon other people's mental model to converge them closer to you own.

That work does not consist just waving at others from over where your own mental model sits, and demanding that they bridge the understanding gap with you from their end, so that your words will make sense to them. Why would they bother? If you want to communicate something, you're the one struck with a motivation to optimize toward a world where other people know the thing you know; the other people who you'd be sharing that information with, have no such motive.

This is why tools like the https://en.wikipedia.org/wiki/Lie-to-children exist: the work of moving someone's mental model closer to your own, of bridging that gap, cannot be accomplished in a single leap; you must often take a circuitous path around (what you consider) the truth, to create a sturdy bridge that will reliably bring people's mental models in line with your own.

Refusal to do this painful-but-O(1) work—presumably under the expectation that the N parties that consume your communication will instead do their own painful mental labor to understand you, for an O(N) aggregate workload—is a disengagement from the entire concept of communication.