I've never been much of a fan of Saussure; in fact, I kind of have a visceral discomfort with the guy. I cannot really articulate why—well, maybe I can but I wouldn't and I shan't. Let's just say that I have technical/philosophical issues with Saussurean interpretation of the nature of language.
I'm a huge fan of John Goldsmith, though. I don't really know if he invented the analytical framework called, Autosegmental Phonology, but he's right up there among thinkers I admire.
Though his 1976 PhD thesis (available here: https://dspace.mit.edu/handle/1721.1/16388) focused on phonology there is a passing section in there about language acquisition that has provided me much food for thought. My main interest in it has to do with dialectal variation, and, along with that, semantic shifting.
In the Inuit language—or, in all languages in general—there is an interesting phenomenon that I call "semantic shift" where the use of a morpheme (basic unit of meaning) stays within the same conceptual class but shifts slightly in meaning. For example, the root:
/qiqi-/ 'to freeze/coldness' in my dialect means "meat that is starting to freeze/frostbite" but in other dialects it means "to feel (extreme) cold". This is what I call "semantic shifting".
This phenomenon is, of course, not unique to the Inuit language. But it usually occurs in terms of historical change within one language, or because of word adaptation (examples in technical and scientific terminologies abound) where the basic sense of the word changes from usage in the source language.
In modern day English (at least, in how Inuit speakers use it), the word /very/ has come to mean "a high degree of (something)" where it once meant "true" etymologically. The King James' bible is replete with examples of how the word was once used:
"Verily, verily I say unto to you..."
All fine and well.
But here I'm focusing on not just the "diachronic" but including the "synchronic" semantic shift within one language family (namely, the Eskimo-Aleut language family). -Diachronic means how language has evolved in historic time and synchronic has to do with where a language (or, more precisely, a dialect) is at a point in time.
Conceptual integrity has been somewhat in the forefront of my mind lately. I'm doing some work on elements of medical terminology—not "translating" but actually talking about medical concepts in Inuktitut, how the derivation and constructive rules of medical terminology work.
Polysynthesis is a very familiar process to Inuit language speakers if not the actual technical term itself. Polysynthesis has to do with compounding different morphemes together to make them grammatical and/or to construct new meaning from its root and affixes.
The additional interesting feature of medical (and, scientific) word elements (and, their constructive rules) is that it draws from more than one source language. For example, the prefixes: bi- and di- (ie, two-ness) may mean the same concept but are used in differing ways:
The prefix /bi-/ (from Latin), for eg, seems to occur where the notion of morphology or tendency is concerned: bilateral symmetry, bicuspid, bisexual, etc.; whereas the prefix /di-/ (from Greek) seems reserved for (deep) structurally-derived two-ness: diphthong (or, a compound vowel), diamelia (absence of two limbs), diandry (an egg fertilized by a sperm with two chromosomes) and digyny (a fertilized egg having two sets of chromosomes from the mother).
(Diachronic) semantic shift is a very powerful concept especially in its deliberately rationalized form (as shown by the usage in medical/physical terminologies). In Inuit-indigenous discourse it may be used not only in standardization efforts but also in modernization efforts which are key to language preservation and enhancement (both drawing from various dialectal forms vulcanized by consesus for specific usage in the different technical fields of discourse).