We know that the modern /u:/ is not from that old /w/ because other words followed the same change - even words without /w/, like “moon”, “poop” (yup) or “boot”. In fact it’s how the digraph ⟨oo⟩ became associated with the sound.
The case of “hwo”→“who” is a bit more complicated. As you said the “wh” digraph used to be “hw”; the change happened in Early Modern times, and it was likely for readability - less sequential short strokes = easier to read. People around those times did other weird stuff like respelling “u” as “o”, as in the word “luue”→“loue” (modern “love”), for the same reason.
However, later on that /hw/ sequence of phonemes started merging into a single sound, [ʍ]: like [w] you round your lips to pronounce it, but like [h] you don’t vibrate your vocal folds. And if that [ʍ] happened before a rounded vowel - like [o:] or similar - it was reanalysed as a plain /h/. So for words like “who”, it’s like the “w” was dropped, just like in “two”, but in a really roundabout way.
And, before non-rounded vowels, that [ʍ] still survives in plenty dialects; for example, “when” as either [ʍɛn] or [wɛn]. This change is recent enough that you still have some speakers in NZ and USA who use [ʍ].
I guess my confusion comes from the fact I’m from a place that only within the last 100 years or so has all but lost the original “hw” in “correct” speech, except in “who”, and I was thinking that the ‘w’ had to have been preserved, especially if the ‘h’ was.
There’s also that the ‘w’ hasn’t vanished in “twenty” (or “twain”, etc.).
No worries - plus the whole thing is damn counter-intuitive, both “sounds fusing together but still conveying two phonemes” and “that sound was analysed as one phoneme, now as another” are kind of weird.
We know that the modern /u:/ is not from that old /w/ because other words followed the same change - even words without /w/, like “moon”, “poop” (yup) or “boot”. In fact it’s how the digraph ⟨oo⟩ became associated with the sound.
The case of “hwo”→“who” is a bit more complicated. As you said the “wh” digraph used to be “hw”; the change happened in Early Modern times, and it was likely for readability - less sequential short strokes = easier to read. People around those times did other weird stuff like respelling “u” as “o”, as in the word “luue”→“loue” (modern “love”), for the same reason.
However, later on that /hw/ sequence of phonemes started merging into a single sound, [ʍ]: like [w] you round your lips to pronounce it, but like [h] you don’t vibrate your vocal folds. And if that [ʍ] happened before a rounded vowel - like [o:] or similar - it was reanalysed as a plain /h/. So for words like “who”, it’s like the “w” was dropped, just like in “two”, but in a really roundabout way.
And, before non-rounded vowels, that [ʍ] still survives in plenty dialects; for example, “when” as either [ʍɛn] or [wɛn]. This change is recent enough that you still have some speakers in NZ and USA who use [ʍ].
I guess my confusion comes from the fact I’m from a place that only within the last 100 years or so has all but lost the original “hw” in “correct” speech, except in “who”, and I was thinking that the ‘w’ had to have been preserved, especially if the ‘h’ was.
There’s also that the ‘w’ hasn’t vanished in “twenty” (or “twain”, etc.).
No worries - plus the whole thing is damn counter-intuitive, both “sounds fusing together but still conveying two phonemes” and “that sound was analysed as one phoneme, now as another” are kind of weird.