Possible Duplicate:
Pronunciation of voiceless alveolar fricative /s/ as ʃ (/sh/) in slang?

My understanding was that the cluster "str", for example in "stress", is usually pronounced /stɹɛs/, just like Wiktionary mentions. Their audio sample confirms this.

However, I sometimes hear a different kind of pronunciation, namely "shtress" or something like /ʃtɹɛs/ (forgive my IPA knowledge).

As an example, Jay Z is rapping about the "New York city shtreets" (here at ~ 0:30), not "streets". The actors from the Jersey Shore seem to do it as well, so I'm starting to believe this is an East Coast / New York thing?

So, regarding this oddity:

  • Where in the English world do people speak like this? Is this even localized to a certain area?
  • Is this a recent development? I didn't hear about this back when I was learning English at school.
  • More formally, is this a "mistake"?

My naive explanation would be that it's simply a natural development because the "sh" forms automatically when you pull back your tongue from forming an "s" to "r".

  • 1
    Don't Scots have a similar, possibly not as pronounced, affectation? I'm thinking of "Shawn" Connery here. P.S. That Jay Z link is pointing to the wrong place. – coleopterist Oct 5 '12 at 13:16
  • 3
    Not an answer, but an expansion: I hear a similar pronunciation amongst northwestern Ontario Ojibway and Oji-Cree first-language speakers, speaking English: "stress" is often pronounced "shtchress," "strange" as "shtchrange," etc. And this morning, my first-language-English, seven-year-old son found it strange that "tree" begins with a T, when, to his ear, it should begin with a CH ("chree"). – JAM Oct 5 '12 at 13:48
  • Interesting. No wonder I didn't find it though – even searching for "shtr" only returns my post. @reg – slhck Oct 5 '12 at 15:17
  • @slhck I wish we had a palatalization tag. In fact I started by tagging your question thus. The problem is, all our questions on palatalization (arguably including yours) have it as the answer. And we are generally very cautious about not tagging questions with their answers. – RegDwigнt Oct 5 '12 at 15:22
  • 2
    I have a hard time regarding "ax" in place of "ask" as an accent. It's a rearrangement of consonants to make a different word. I mean, the person knows how to say "k" and how to say "s", but is just mixing their order. – Kaz Oct 5 '12 at 18:37

After digging for a while, I found the answer in a paper called Assimilation at a Distance by Wayne P. Lawrence.1 It's a little dated (2000), but also cites studies from 1993 and earlier describing this phenomenon.

The recent history of American English includes a sound change that seems to have gone unattested in the scholarly literature.1 This is the change of /s/ to /ʃ/ before /tr/ (i.e., a PHONEMIC CHANGE), which involves a palatalization of the initial sound in the cluster /str/, typically in initial position but not exclusively.

The author further mentions that "this phonemic change seems to be neither dialectal nor regional." However, it's definitely an American English peculiarity. It seems that it was first heard on television or radio broadcasts and became popular over these channels. Of course, the internet helps spreading such a linguistic "oddity" – and even I know I picked it up from television shows and continue to (subconciously) use this pronunciation.

Regarding the phonetic principles behind this change, Lawrence summarizes:

More concretely, this pronunciation is an implicit "assertion" on their part that in their grammar /r/, being [+ interrupted], is marked with respect to the opposition INTERRUPTED versus CONTINUOUS - exactly the same way that [ʃ], being [+ compact], is marked with respect to the opposition COMPACT versus DIFFUSE. The marked character of /r/ is prompted and confirmed phonetically, of course, by its high degree of retroflection in American English.7

This would explain the fact that we only hear this in American English. What's interesting is that the same does not appear to happen to "spr", which can be explained by the phonetic differences in "t" and "p":

I would surmise, therefore, that /str/ lends itself to realization as [ʃtr] while /spr/ remains unaltered also due to the phonological value of /t/ and /p/.

Another more recent paper was written by David Durian.2 It goes into the details of social background and "shtr" as well as providing some more explanations for its development.

1 – Lawrence, Wayne P. (2000) "Assimilation at a Distance," American Speech Vol. 75: Iss. 1: 82-87; doi:10.1215/00031283-75-1-82
2 – Durian, David (2007) "Getting [ʃ]tronger Every Day?: More on Urbanization and the Socio-geographic Diffusion of (str) in Columbus, OH," University of Pennsylvania Working Papers in Linguistics: Vol. 13: Iss. 2, Article 6

| improve this answer | |
  • A very informed comment, far beyond my areas of expertise. But I know that when I do this, I am copying the laid-back tones of Humphrey Bogart from the move Casablanca. It seems to me that it is related to an exaggerated casualness, deliberately using laziness of mouth to convey a sense of control over a situation. "Shtreets" and "Shweetheart" are done in similar vein. Any comments on this idea would be appreciated. – shipr Oct 5 '12 at 15:20
  • It is encouraged to accept your own answers on the SE network, in case you don't know ;). – terdon Jun 9 '13 at 20:43

The phenomenon has been studied by sociolinguists, beginning with Labov. This isn’t my area, so I can’t give you up-to-date references.

The same, or a similar, phenomenon is found in Sydney, Australia, where I grew up. I hasten to add though, the pronunciation isn’t /ʃtɹɛs/(*) , but /çtɹɛs/: i.e., a voiceless palatal fricative (as in German ich), not a voiceless palato-alveolar fricative (as in English wish).

There are certainly good articulatory foundations for this behaviour, along the lines you suggest, to do with preparing the tongue to produce sounds in the same (articulatory) vicinity. (For which reason, /s/ tends to be preserved in /spr/ and, especially, /skr/: the tongue does not stay near the alveolar ridge in these cases.)

Whether this is a “mistake” depends on how normative you are. The Queen doesn’t speak that way. If everything the Queen doesn’t do is a mistake, then this is too. Otherwise, it’s just another legitimate form of dialect variation.

(*) I assume your omission of /t/ here is accidental, given its non-omission in the title of the question.

| improve this answer | |
  • Yes, thanks for pointing out the mistake – I simply forgot a t there. – slhck Oct 5 '12 at 14:19
  • +1. Some more-up-to-date references are mentioned in this blog post by Neal Whitman; he also gives some alternative theories. – ruakh Oct 5 '12 at 15:50

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