Linguists dumb as usual
Linguists tend to be accurate and clever observers and recorders of details. Linguists tend to be really dumb analysts. They constantly reach absurd conclusions.

Fine example here.

Summary: A unique hand-sign language has been observed in one Arab village in Israeli territory. This village has a lot of congenitally deaf people, and they have developed their own SL without any contact with existing mature SL's. Linguists noted that this language consists almost entirely of 'pure meaning'; each hand sign is a literal description of an object or action. It hasn't developed any tendency to economize on basic forms by combining or prefixing.

Reasonable conclusion: It's a young language in a restricted and simple environment. Pretty much the same thing as private twin-lingua. It hasn't spread to any outside adult learners so it hasn't needed any sort of rules or formalization.

Linguist conclusion:

To linguists, this is akin to finding a planet on which matter is made up of molecules that don’t decompose into atoms.

It’s not hard to imagine that things could have been otherwise. In principle, we could have a language in which sounds relate holistically to their meanings—a high-pitched yowl might mean “finger,” a guttural purr might mean “dark,” a yodel might mean “broccoli,” and so on. But there are stark advantages to duality of patterning. Try inventing a lexicon of tens of thousands of distinct noises, all of which are easily distinguished, and you will probably find yourself wishing you could simply re-use a few snippets of sound in varying arrangements.

As noted by Elizabeth Svoboda in the current issue of Nautilus, the dominant thinking until fairly recently was that universal linguistic properties reflect genetic predispositions. Under this view, duality of patterning is much like an opposable thumb: It evolved within our species because it was advantageous, and now exists as part of our genetic heritage. We are born expecting language to have duality of patterning.

You've got it backwards as usual. Even in sound-based languages, we aren't born expecting this atom/molecule structure. Babies expect every object AS CLASSIFIED BY BABIES to have its own symbol. As a kid grows beyond age 2, he slowly picks up a more refined CLASSIFICATION OF OBJECTS, and along with this refinement picks up the more refined set of terms.

The degree of atomization is a variable that depends on historical and geographical factors, not on linguistic basics. Compare English with German. English has a much larger set of unbreakable atoms because we decided to grab new words from elsewhere. Many of these unbreakable atoms were compounds in Greek or Latin, but because normal speakers don't know Greek and Latin the compound is felt as an indissoluble whole. German decided to keep a strictly limited set of native atoms, building new meanings with multiple layers of native prefixes and suffixes.

Linguists are making the same mistake that mathematicians make. Mathies believe falsely that all of math is 'built' from the atoms of set theory and number theory, because mathies have learned to analyze math into those atoms. In fact math was built from a natural sense of proportions and comparisons. Logarithms and percents are the real starting point. Set theory is an EXTREMELY recent development within an EXTREMELY NARROW set of weirdos, and has no connection to the way cultures or individuals develop math skills.

= = = = =

Sidenote: I didn't intend it, but this item is directly parallel to the previous item, about an idiot writer who mistakenly believes that "understandable writing" requires breaking down words into minimal native atoms.
  ¶ 3:55 PM