2

This question was inspired by the interesting discussion here: Why isn't the T in "relative" flapped?

It seems like the adverb already and the two-word phrase all ready should be pronounced differently, but as far as I can tell, both sound exactly the same.

For comparison, consider the other phrase/adverb pair all ready and already, where there is a difference:

"The suitcases are all ready"

doesn't sound the same as

"The suitcases are already..."

because the stress pattern is different ([all ready] as opposed to [al ready]). Additionally, the "l" sound at the end of "all" in the phrase goes on for longer than the "l" sound in the adverb, corresponding to the separation between the two words. If someone said the second sentence (fragment) out loud, it would leave the listener asking "already what?"

In contrast, I wouldn't be able to hear the difference if someone incorrectly substituted the adverb altogether for the phrase all together. They both have the same sequence of stressed and unstressed syllables ([all to ge ther] vs. [al to ge ther]). More surprisingly, the adverb also keeps the aspirated "t" sound found in "together," even though that t is between a stressed and an unstressed syllable. This seems very exceptional, since that context usually requires the substitution of the "flapped t" for the "aspirated t." But, if this were the case, then "altogether" could be spelled "aldogether" without changing its pronunciation, which it can't. ("Aldogether" just sounds like I have a stuffy nose.)

This also contrasts with "relative"/"reladive" like in this question, where the two pronunciations are interchangeable (the flapped t one being more common when it's said faster). "Relative" is also not an exception to the rule cited in the accepted answer, since there the t comes between two unstressed (or at least only tertiary stressed) syllables. The cited paper states that flapping the t is optional in that case.

I'm interested in what the context (either historical/usage-related or some less well known phonological rules, or something else?) could be behind the pair not being differentiable by sound, and only by spelling, but none of the answers to the other question mention the word "altogether." No one seems to have asked this specific question yet, as this one is about meaning and not pronunciation.

Here is a reference for the pronunciation of "altogether," confirming it always has a distinct "t" and not a flapped t/d sound: https://en.wiktionary.org/wiki/altogether. This proves that I haven't just been mishearing and mispronouncing the word the entire time!

alphabet
  • 18,217
Quack E. Duck
  • 285
  • 2
  • 2
  • 12
  • 2
    N=1, but I do flap the /t/ in altogether, and don't aspirate it. – alphabet Nov 27 '23 at 07:29
  • 1
    I could be misunderstanding your question, but according to https://en.wikipedia.org/wiki/Flapping it largely depends on which version of English you are speaking, American, London, RP etc. I speak Southern/London English fairly close (I think!) to RP and would fully pronounce the 't' in both cases. I'm minded of the two phrases "All together now, sing after me …" and the Danny Kaye song from Hans Christian Anderson "The King is in the Altogether", both of which I hear as unflapped t. – Peter Jennings Nov 27 '23 at 13:29
  • 1
    It should be noted that Wiktionary's IPA transcriptions, like those used by many dictionaries, do not consistently show AmE flapping. So that's not a reliable source for whether or not most speakers flap that /t/. – alphabet Nov 27 '23 at 17:15
  • 1
    I also suggest editing the title to remove the phrase "minimal pair" and just ask about homophony, since this question is only tangentially about phonemic analysis. – alphabet Nov 27 '23 at 17:16
  • @alphabet OK, did it – Quack E. Duck Nov 27 '23 at 17:19
  • @PeterJennings I may be misunderstanding your comment :) but it isn't surprising to me that the t would be pronounced in a variant that's known for always pronouncing t's. The surprising part is that it's pronounced (or at least, I believe it usually is, although after alphabet's comment I'm less sure) in a variant that doesn't usually keep the aspirated version in that context. That's why I added the [tag:american-english] tag specifically. I generally always/almost always use the flapped sound when it's possible, but not in the word altogether. – Quack E. Duck Nov 27 '23 at 17:25
  • 1
    Incidentally: for me, altogether has flapping, but all together also often has flapping, which can cross word boundaries. – alphabet Nov 27 '23 at 17:26
  • 1
    The answer is presumably the presence of a syllable/morpheme boundary between the /l/ and the /t/ in altogether, which could like a word boundary in terms of its influence on flapping. But I'm not sure of the details. – alphabet Nov 27 '23 at 17:33
  • 1
    Note that there's also no prefortis clipping in the first syllable of altogether; I think this applies to people who don't flap that /t/ also. In other words: the first syllable of altogether has the same vowel length as all, not the shorter one in salt. (Compare the length of the first vowel in keychain to the one in teaching--the morpheme boundary in the former stops the vowel from being clipped.) – alphabet Nov 27 '23 at 17:38
  • Maybe it's relevant that there's no actual vowel sound between the "t" and the "g"? – Quack E. Duck Feb 25 '24 at 03:23
  • The 'release' phase of the 't' sound kind of fills in for the schwa vowel there would normally be in an unstressed syllable. Is there a name for this? As far as I know only 'sonorant' consonants like 'n' can be syllabic. – Quack E. Duck Feb 25 '24 at 03:26

0 Answers0