Talk:English relative words

That which

When what is the relative word (in such fused relatives as I paid a lot for what I bought and She ignored what I'd told her and What he said appalled me), it can be replaced by that which. (Yes, the formality may make it sound incongruous, but I hold that it's fully grammatical.)

One interpretation that might come to mind (not least because that is stressed) is that the result is no longer a fused relative and that it's instead a regular relative clause, and the antecedent is the DP (and not the subordinator) that. But if distal that is possible, I'd expect proximal this to be possible too. And yes it might be, but whereas the versions with that which make integrated RP intonation far more likely, the versions with this which seem to force supplementary RP intonation (likely written as, for example, She ignored this, which I'd told her, and straightforwardly contrastable with distal She ignored that, which I'd told her).

And so I'd say that in She ignored that which I'd told her, we still have a fused relative. Is that which a compound relative word (one that, like no one, is written as two words merely as a matter of convention), or what's going on here? (Should we perhaps abandon the notion that that is a "relative word"?) -- Hoary (talk) 00:18, 18 November 2023 (UTC)

Lovely observation, but I think you're right that that that is a determiner, and not a subordinator. "The underlined sequence here is an NP, not a clause; it is distributionally and semantically comparable to expressions that are more transparently NPs, such as the money which you gave me yesterday or the very formal that which you gave me yesterday" (CGEL, p. 65).--Brett (talk) 02:07, 18 November 2023 (UTC)

That's good: less additional work to be done! (Yet to be explored: wherein and the like.) -- Hoary (talk) 11:26, 18 November 2023 (UTC)

/hw/

If (or so far as) "the historical /hw/ sound" really was "a sound" (a discrete phoneme), then -- quotations aside, of course -- shouldn't it be written either (A) as /ʍ/ or (B) with a tie bar?

Unsure of the relative merits of (A) and (B), I haven't "been bold". -- Hoary (talk) 08:33, 19 November 2023 (UTC)

I was bold. What do you think?--Brett (talk) 20:19, 19 November 2023 (UTC)

Um, whither the bolditude, Brett? Sorry, I'm still lost. Perhaps I wasn't clear. As I understand or misunderstand it:

• ⟨/hw/⟩ represents a pair of phonemes: /h/ immediately followed by /w/.
• ⟨/h͡w/⟩ represents a single phoneme, indivisible in the language described.
• ⟨[ʍ]⟩ and ⟨[h͡w]⟩ represent the same sound.

I wouldn't be at all surprised to learn that I'm mistaken. But if I'm not mistaken, then (unless we're quoting, etc), when we're describing English previous to (or a lect that's unaffected by) the whine–wine merge, shouldn't we avoid ⟨/hw/⟩ and instead plump for either one of ⟨/ʍ/⟩ and ⟨/h͡w/⟩? -- Hoary (talk) 11:56, 20 November 2023 (UTC)

Perhaps the answer is, yes, contemporarily, and no historically? Brett (talk) 13:54, 20 November 2023 (UTC)