A recent ELL HNQ post about "sure" put a question in my head: Wait, why is "sure" pronounced the way it is?
The initial consonant is a voiceless postalveolar fricative [ʃ] as opposed to a voiceless alveolar fricative [s]. Another common word example is "sugar". I have always considered the phonetic oddity of "sugar" easily explainable. It originally came from Sanskrit by way of Arabic and the "sh" sound as a holdover indicating its provenance makes sense. Sugar originated in the Indian subcontinent and for much of European history it had to be imported into Europe by Middle Eastern merchants. Etymology Online:
sugar (n.)
late 13c., sugre, from Old French sucre "sugar" (12c.), from Medieval Latin succarum, from Arabic sukkar, from Persian shakar, from Sanskrit sharkara "ground or candied sugar," originally "grit, gravel" (cognate with Greek kroke "pebble").
That Etymology Online page has one vague line on its pronunciation change that doesn't tell us much besides unsubstantiated speculation:
The pronunciation shift from s- to sh- is probably from the initial long vowel sound syu- (as in sure).
But what about "sure"? It doesn't seem to have come from either Arabic or Indian languages
sure (adj.)
early 13c., "safe against attack, secure," later "firm, reliable" (c. 1300); "mentally certain, confident" (mid-14c.); "firm, strong, resolute" (c. 1400), from Old French seur, sur "safe, secure; undoubted, dependable, trustworthy" (12c.), from Latin securus "free from care, untroubled, heedless, safe").
Another word with a similar pronunciation pattern I can think of is "sumac", but in addition to [ʃ] it can also be pronounced with [s]. But we don't pronounce "super" with [ʃ], nor "suicide", nor "sue".
