As I might have mentioned in previous postings to this section, I've studied phonosemantics (the connection between linguistic form and meaning) for 40 years now, and have examined around 150 languages from around the world looking for patterns, and trying to analyze them where they are present.
Of these languages, Santali (a Munda language from northeastern India, a member of the larger Austroasiatic stock most of whose members are in Southeast Asia (Mon-Khmer), seems to be the richest in the most transparent words, called expressives (though ideophones are more commonly used to name such words in languages in other regions of the world, as in Africa).
A good (though still minor) portion of the entire lexicon of Santali consists of such transparent words. And because of peculiarities of the North Munda languages (ambivalent (or 'flexible') word-class assignment, for example, most of these words can be used either expressively, adverbially, adjectivally, verbally, or nominally, just by attaching the right morphology to them and plonking them down into the right positions in a sentence.
Anyway, for the past several weeks I've been collecting the expressive words that contain the same vowel skeletons. The canon is normally ©V(N)CV©, and can be reduplicated, either in whole or in part, and either vowel or consonant may be shifted to another in the reduplicated part in specific patterns. I've been working so far on fully reduplicated expressive words where the vowels and consonants remain the same in both halves.
Thus far this pattern has begun to emerge. Having completed collecting such forms where the vowels were all /i/, /u/, and /e/ (the latter both tense and lax variants), /i/ strongly associated with the idea of passage of small amounts of material through a very narrow channel, collecting at the release aperture (the way a droplet of moisture would form, for example), and then relatively energetic release from that latter position by adding force (like when you flick something away). /u/, on the other hand, associates strongly with the idea of collecting INSIDE an enclosed space such that immediate escape is impossible (as when you fill a balloon with water, or a developing pregnancy). Sometimes these notions are highly abstract, and apply to domains outside fluid flow- as when one walks, one might stub your toe against a stone stopping you (at least for a time) from proceding.
I've also completed collecting the terms with the two variants of /e/, and THEY strongly associate with release from a channel as well, but not exactly the same as with /i/. Instead, the exit aperture of the channel spreads, so that the fluid flow slows considerably and its height falls (as in a river when it hits a delta region). I'm currently working on those expressives with /o/ (again both tense and lax). These seem to share a lot with the /u/ set, paralleling the behavior of the /e/'s versus /i/. Here the flow is again rebuffed, but not for long- the moving fluid manages to escape the barrier or impedience, either finding a way around or over, under, or through it. The latter sense seems to be the more common one, leading me to suspect that this is the opposite of the widening channel notion. If the channel NARROWS, then the speed of the fluid will increase. Bernoulli's law or something similar. One sees similar types of behavior with tsunamis, where as the shoreline approaches and the depth of the water decreases, the height of the incoming wave increases dramatically, allowing it to overtop any potential barriers (though I've heard the wave slows considerably rather than speeding up).
I'll predict (but don't make any bets just yet) that the expressives in /a/ will present the notion of freedom of flow with minimal impedence. This seems to be the crosslinguistic pattern, with /a/'s associating with flat, level extents of surface, unbounded.
And as for other languages, I remember that Tzotzil Mayan expressive verbs (there's a technical name for them I'm having trouble remembering) have a very similar spread of senses with different vowels specified.
Perhaps this is a kind of default iconicity? Based on the actual shapes of vowels in the mouth and how they modify air flow?