Tuesday, 15 November 2011

that'd be OK

A simple question from a Japanese university student: how is that’d pronounced?

My immediate answer was to tell him that it’s ˈðæt əd. I still think that’s the riɡht short answer, but things are actually a little more complicated.

Some relevant variables:
(i) The that element will have a strong vowel, ðæt, only if it is demonstrative, as in that’d be fun, that’d be OK, I don’t think that’d work. If it is a relative pronoun, as in someone that’d been here before, it will almost always be weak, ðət.
(ii) The ’d element may stand not only for would, as in the examples given, but also possibly for had, as in that’d never worked in the past. This makes no difference to the pronunciation.
(iii) There is also the possibility of pronouncing that’d as a monosyllable, ðæd or maybe ðæt, perhaps with further contextual assimilation: that’d be OK ˈðæbbi əʊˈkeɪ.
(iv) for speakers who use t-voicing (esp. AmE), the intervocalic stop/tap in the disyllabic version will be voiced. Here’s a case in point from YouTube.

I can’t find any dictionary entry for that’d that includes pronunciation. Many dictionaries do not even have entries for that’ll and that’s, but leave their pronunciation to be inferred from entries at that and (if you are lucky) ’ll and ’s. LPD does have entries for these contracted forms, though.
The LPD entry at ’d mentions the comparable it’d but not that’d. But now I wonder if I ought to add the possible monosyllabic versions of both.
More generally, in what varieties and styles is it usual to write contracted that’d rather than the usual full that would, that had? I’m really not sure.


  1. I think that if I reduce the that as well as the would in that had been there before, I feel impelled to assimilate to ðətəbbɪn 'ðɛə bɪ'fɔ: — and probably reduce to ðətəbɪn.

  2. David:
    Would you reduce that phrase to ðəbbɪn 'ðɛə... (following the ˈðæbbi əʊˈkeɪ example mentioned by John) or would it be necessary for the initial "that" to be stressed?

  3. A similar example is that've etc. I notice that native speakers often write "of" rather than "'ve", at least in web forums.

    And in what styles is it usual to write contracted forms? It's probably to easy to just say less formal than more formal. It's also probably indvidual. Forums, emails, SMS text, private letters. Less private letters? Depends on you and who you're writing to. For publication? The editor would have the last word.

  4. Interesting -- after doing a bit of thinking, I'm reasonably sure I always say ðæd for a contraction of "that would", but never for "that had". In the latter case I think it's ˈðæt əd, even though I generally flap in such contexts. Perhaps it's even ˈðæt həd; I don't think I like to contract "that had" like I do "that would".

  5. I associate the monosyllabic versions with AmE; especially "It'd be nice if...". Just an impression.

  6. So, "you'd", "he'd", "she'd", "we'd" are pronounced [ju:d], [hi:d], [ʃi:d], [wi:d].

    But, couldn't they be (also) [jʊd], [hId], [ʃId], [wId]?

    I'm not a native speaker, but I normally say, for example, "[jʊd bI] surprised", "[hId] gone away", etc. Am I wrong?

    In other words, can't both the pronoun and the verb be reduced (weak)?

    More generally speaking, I personally tend to use two, three and even four or five weak forms in succession ("I know [ðət hId bI] right", or "[ðət hId bI ðə] winner"). What do native speakers do?

  7. Wiseman: yes, of course. If you looked up "you'd" etc in LPD, where they have their own entries, you'd find the weak forms. (Incidentally, I naturally write them in the modern way as jud etc rather than jʊd, since the weak forms of these pronouns have the happY and thankyOU vowels rather than KIT and FOOT as such.)

    There is no problem with successive weak forms.

  8. PS: Kensuke Nanjo points out that the Genius English-Japanese Dictionary, fourth edition (2006), of which he is the Phonetics Editor, records both a strong form and a weak form of "that'd", in each case with a schwa between /t/ and /d/.