What’s strange is that many other languages don’t allow words starting with pt- or ps- neither but have no trouble pronouncing Greek loan words that do.
I have a private theory about that, actually (that is, not backed up by research yet to my knowledge).
I think this is due to accidental gaps, that some languages allow for clusters that just don’t happen to appear in those languages by an accident of history (e.g. they allowed them at one point but they were eliminated by a phonotactic filter that no longer exists in the language, etc.), so when they borrow a word with that string now, they can pronounce it no problem.
If you think about phonotactic constraints as being the result of constant rankings (as in models like Optimality Theory), this should even be predicted as a form of Emergence of the Unmarked (though stop clusters are pretty marked, so this would be more like “local” or “coincidental” unmarkedness).
I also think that studying borrowing adaptations like this would give us a more accurate picture of the overall constraint ranking of a given language than just restricting ourselves to native words.
Unless you’re talking about Scots, the closest languages to English are separated by at minimum more than a thousand years, which is plenty of time for those constraints to change significantly.
I’d even expect different dialects of English to behave differently when adapting loanwords, because they already show plenty of phonotactic differentiation.
What’s strange is that many other languages don’t allow words starting with pt- or ps- neither but have no trouble pronouncing Greek loan words that do.
I have a private theory about that, actually (that is, not backed up by research yet to my knowledge).
I think this is due to accidental gaps, that some languages allow for clusters that just don’t happen to appear in those languages by an accident of history (e.g. they allowed them at one point but they were eliminated by a phonotactic filter that no longer exists in the language, etc.), so when they borrow a word with that string now, they can pronounce it no problem.
If you think about phonotactic constraints as being the result of constant rankings (as in models like Optimality Theory), this should even be predicted as a form of Emergence of the Unmarked (though stop clusters are pretty marked, so this would be more like “local” or “coincidental” unmarkedness).
I also think that studying borrowing adaptations like this would give us a more accurate picture of the overall constraint ranking of a given language than just restricting ourselves to native words.
The languages I’m most familiar with are quite closely related to English, so I don’t think that’s really the case.
Actually, one thing I can think of is that English trends to aspirate initial stops, which probably makes those clusters harder to pronounce
Unless you’re talking about Scots, the closest languages to English are separated by at minimum more than a thousand years, which is plenty of time for those constraints to change significantly.
I’d even expect different dialects of English to behave differently when adapting loanwords, because they already show plenty of phonotactic differentiation.