8

Why is the 't' pronounced like 'ch' sometimes or even like "t+sh"? Do the English phonotactics allow for a word to start with "ch+r" ('ch' as in 'chair', not as in 'Christmas')?

This doesn't appear to be true for all languages. The 't' (three) in the Croatian word 'tri' is pronounced quite differently from the 't' in the English word 'tree', even though they are usually pronounced the same. Also, the word for 'black' in the Kaikavian dialect of Croatian is 'črn' ('č' representing a 'ch' sound), while the word for 'thorn' is 'trn', and those are pronounced quite differently.

  • 2
    I don't know what the closevotes are for, but you might have better luck if you include what accent you're observing here. Make sure your question is still about linguistics though. – OmarL Dec 23 '17 at 11:12
  • 4
    This is simple co-articulation: English /r/ is retroflex in many dialects, and presumably all of the dialects where the phonomenon you describe occurs. In Swedish for instance, where /r/ can be similarly retroflex, co-articulation causes the sequence /rs/ to be pronounced [ʂ]. Because it's not phonemically relevant at this point, but just a phonetic realization of the /tr/ sequence, phonotactics do not enter into play: even if /t͡ʃr/ is not allowed by English phonotactics, [t͡ʃɹ] can still be as a realization of /tr/. – LjL Dec 23 '17 at 15:58
4

This phenomenon occurs in many varieties of English, but doesn't got as for as neutralising the contrast. The transcription that @Wilson gives ([t͡ʃɹi]) captures this reasonably well. This might be a consequence of the backward movement of the tongue after [t] during which it passes the place of articulation of [ʃ].

As such, the phenomenon is one of co-articulation and not a phonological process. So an onset /t͡ʃɹ/ (which might be spelled chr if it weren't for the confusion with /kɹ/ as in Christmas) is still illegal in English phonotactics. However, over time this co-articulatory process may become so common that it is phonologised and the contrast is neutralised.

In fact, the eminent phonetician John Wells argues in his blog that it already has been (thanks to @Sumelic for the link). His case in point is the pair century - sentry.

| improve this answer | |
  • So you think that phonology deals only with phonemics or distinctive changes? This is a peculiar view. – Greg Lee Dec 23 '17 at 17:17
  • 1
    I think there may be some speakers who have at least a perceptual neutralization between words like "century" and "sentry". See JHJ's and malti's comments beneath the following blog post by John Wells: how do we pronounce train ? – brass tacks Dec 23 '17 at 20:36
  • 2
    While it's not entirely clear that the disyllabic pronunciation of "century" should be transcribed with /t͡ʃɹ/, that at least seems to be the most obvious transcription given the fact that the trisyllabic pronunciation standardly has /t͡ʃ/ rather than /t/. – brass tacks Dec 23 '17 at 20:42
  • @GregLee do you have a different analysis of this matter? – robert Dec 25 '17 at 3:34
  • 1
    The test is hard to perform since it depends on the sci-fi "loanword" chrowl and the obscure language name Chrau, but the former could neutralize with trowel. – user6726 Dec 25 '17 at 19:57
3

The ch-like sound in "tree" is a phonetic change, not a phonemic change, so the resulting sounds are not adapted to the phoneme system of English. It is due to an assimilation of obstruency, in my view (I haven't seen anyone else describe it this way). The coronal r becomes an obstruent after a coronal obstruent when both sounds are in the onset of the same syllable. (It is very similar to the change of y (yod) to sh in "tune", which is also an assimilation of obstruency and voicing.)

There is also assimilation of voice, so after d, in "dream" e.g., the r remains a voiced fricative, but devoices in "trip". Since r before a stressed vowel in the same syllable is generally rounded, that rounding remains in "trip/dream". It may be that some retroflexion of "r" is also carried over to the tr/dr obstruent.

My formulation implies that r in words like "shriek", "shrike", "ashram" should also become an obstruent. I'm not sure that happens. Unlike the change of y to an obstruent, the change of r does not happen when r and the preceding coronal obstruent are in different words.

| improve this answer | |
-3

Your question is based on a false premise. I do not think there are any speakers of English who do not distinguish between "trees" and "cheese", or between "train" and "chain".

| improve this answer | |
  • 3
    I don't claim they don't distinguish between "trees" and "cheese", I am claiming that "tree" is pronounced as if it was spelled "chree" or "tshree". Don't you agree? Try pronouncing it. – FlatAssembler Dec 23 '17 at 11:30
  • Many North American speakers pronounce the "tr" in "tree" with the tongue slightly curled back. This is not the case with the "ch" in "cheese". – fdb Dec 23 '17 at 11:42
  • 1
    @fdb I absolutely disagree with this. (At least some) Caribbeans pronounce "truck" the same as "chuck". In Jamaican Patois they're pronounced the same. Anyway Flatassembler wasn't asking about that. In my native BrE, tree really may be pronounced as t͡ʃɹiː. You may check Wiktionary for this and other pronunciations. – OmarL Dec 23 '17 at 12:48
  • Yes, this occurs in many varieties of English, but doesn't got as for as neutralising the contrast. The transcription that @Wilson gives captures this reasonably well. My hunch is that this is a consequence of the backward movement of the tongue after [t] during which it passes the place of articulation of [ʃ]. – robert Dec 23 '17 at 14:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.