Phonological Resistance
Not so much a very insightful post as a mere musing.
It has always puzzled me how one language can be so resilient to changing their phonologies when loaning words while others aren't.
There's languages that take on whole new phonemes purely from loaning from other languages. English Zoo has a word initial /z/ which in native words is illegal. Dutch goes even further, all instances of /s/ and /f/ are loan phonemes. It's highly dependent on the region in how much they merge with /z/ and /v/, but in my speech I have a clear distinction between fee 'fairy' and vee 'lifestock'. Other languages have taken on completely new illegal clusters in their language like āyiskrīm in Arabic with a CCC cluster. There's also some initial CR words though I can't think of them right now. Nevertheless Arabic always stayed quite resistant, often adapting loanwords to known consonant patterns.
bank 'bank' has a plural bunūk.
Other languages have gone even further completely taking over larger parts of illegal phonological traits. Swahili is probably a good example. stampu 'stamp' with a previously illegal st cluster at the beginning of a word.
It also amuses me that french has a word like scolaire after trying so hard to get rid of that cluster over the centuries creating école out of Latin schola.
And when you see languages this open to changes you wonder why Japanese after all this time and such an enormous influx of loanwords from english still refuses to pronounce love any different than /rabu/. Why, although often written as such, video is still pronounced /bideo/ and star trek is still /sutaatorekku/. Why is Japanese so resistant to changes in phonology and why aren't other languages?
The 'contact' or 'influence' argument gets you a long way, but doesn't fully cover the problems it brings. How can Latin be said to have such contact with French that it would influence their phonology, it's a dead language after all, that wasn't spoken widely either.
And maybe something could be said about English influence in Egyptian Arabic, but is such influence stronger than anglo-mania that currently spreads throughout Japanese? Not to mention the rather great, though not as great resilience to influence in Chinese loans as well. Sure it introduced some Cy clusters, but still Chinese loans sound, and never did, nothing like Chinese.
Maybe it has something to do with how limiting a phonology is? That would stand in the way why Swahili didn't follow Japanese's path. Maybe it is actually related to the orthography too?
Any musings are welcome.
Comments
For the substitutions then, [v] being an alien sound in the area, it's no wonder they substitute /b/, and similarly, since their /r/ can be partially lateral, it therefore matches both "usual" /r/ and /l/. In Finnish, we similarly match /v/ and /w/ into /ʋ/. I was in my twenties before lerning that there even [i]is[/i] as distinction between [v] and [ʋ].
I don't see the re-introduction of sC- in French as very problematic either; AFAIK it's found mostly in international words (very recent loans aside), and speakers familiar with those should also be familiar with a number of languages (Latin or not) that do have such clusters. So we are not talking about Latin affecting French as a whole, as much as every non-Western-Romance IE language of Europe affecting the French of the academically literate, and that then diffusing to affect the rest of French.
Finally, what do you mean "all instances of /s/ are loan phonemes in Dutch"?? Is s before a consonant never /s/?
So my point was, that languages resistant to changes in phonology, can't be readily explained. There's good motivations. The problem is that these motivations usually don't go for 'all' languages. Why does phillipino struggle less with the loaned /f/ than Korean? Korean absolutely refuses to introduce /f/ and continues to use /p/, while at least in some part of the speaking population of the phillipines /f/ really does exist.
What seems like a reasonable explanation for resisting change in one languages just refuses to transfer to another language with a comparable phonology. That's odd, shouldn't there at least be some kind of study on why some languages resist harder than others?
As for /s/ in Dutch, s before a consonant is an allophone of /z/. You can't say [z] is an allophone of /s/ because then minimal pairs of fee and vee would get in the way. Though of course for synchronical discriptions of the language the issue is hardly relevant. I bet plenty of phonologies of Dutch would say that /s/ as in soep is the same /s/ as in schrijven, but diachronically it's wrong. Depends on how you look at it I guess.
This is a good topic which I admit to not thinking about deeply until you've mentioned it now. When I was reading this entry, my mind immediately thought of the curious instance of Etruscan streteθ written in the Liber Linteus. I call it "curious" because all other instances of word-initial clusters I've collected into my database have at most two consonants, never three. So this fact has already suggested to me that the word is a recent loan, probably from Latin. Etruscan then would have been adopting the occasional three-consonant cluster from neighbouring Indo-European languages.