The next time you hear someone pronounce "street" as "schtreet," you'll know why.
Reading note: Read /ʧ/ as “chuh,” and /ʃ/ as “shuh.”
Reading note: Read /i/ as a long E as in see, and /I/ as a short I as in sit.
I need to start this next segment with a warning. I’m going to talk about something that you may never have noticed, but after listening to this segment, you won’t be able to not notice it. Are you ready? OK. Listen to how I pronounce this next sentence: I painted a shtripe shtraight down the middle of the shtreet, then shtruck a pose as I shtraddled the line.
Some People Move the S Back
Did you catch it? I pronounced the all words that began with an STR consonant cluster as if they actually began with SH. Instead of stripe, street, straight, struck, and straddled, I said shtripe, shtreet, shtraight, shtruck, and shtraddled.
Linguists call this pronunciation S-retraction or S-backing, because the SH sound is made with the tongue slightly farther back inside the mouth than it is for the S sound. So you’re moving the S farther back in your mouth: S-backing. When I first read about it, I didn’t know what the author was talking about, but pretty soon I heard it for myself. For a while I was keeping track of the various people I heard, but there are too many for me to keep track of now.
Is It a Mispronunciation?
Some speakers consider this a mispronunciation, given that the words are spelled STR, not SHTR. That’s true enough, but let’s widen our focus a little bit. Let’s look at words that start with just TR, like truck, trap, and transmogrify. If you listen carefully, you can hear that for many speakers, the T is actually pronounced more like a CH sound.
Listen again: Instead of saying [tr]uck, [tr]ap, and [tr]ansmogrify, I’m saying [ʧr]uck, [ʧr]ap, and [ʧr]ansmogrify. Kids who are just learning to spell will sometimes even write these words with a CH instead of a T. Except for transmogrify—by the time they’re trying to spell that, their spelling skills are probably advanced. The point is, nobody calls this a mispronunciation. If they notice it at all, they figure it’s just how T is pronounced when it comes before an R. In the same way, for speakers who say shtripe and shtreet, this is just how S is pronounced when it comes before TR.
Pronunciation Changes Are Common
Changes in pronunciation like these happen all the time in every spoken language. In fact, sound changes like this one are the kind of language change that has been studied the most intensively and for the longest time. In the old days, we could only study sound changes on dead languages, such as the changes that Latin underwent as it evolved into the modern Romance languages, or changes that happened to archaic Germanic languages as they evolved into modern ones such as English, German, and Swedish. Nowadays, though, we can observe sound changes as they happen. If you’ve encountered S-backing, you’ve been hearing language history in the making.
Why Does S-Backing Happen?
Two of the most fundamental questions about these sound changes are why they happen in the first place, and how they spread to new speakers. So why does S-backing happen? Several linguists have offered opinions. One possibility is that it’s a continuation of the same sound change that turns those T’s into chuh /ʧ/ sounds in words like truck. The S is turning into shuh /ʃ/ to be more similar to the /ʧ/ that comes right after it, a process called assimilation.
However, linguist Michael Shapiro argues against this idea. He points out that the T-to-/ʧ/rule that operates on TR clusters, like those in the word truck, has an exception: It doesn’t happen in STR clusters. If I put an S in front of truck to get struck, I pronounce it as [str]uck, not as s-ch-truck [sʧr]uck. This is true for many S-retractors, too, who pronounce it as shtruck [ʃtr]uck, not as shchtruck [ʃʧr]uck, a fact noted by linguists Richard Janda & Brian Joseph. Shapiro’s solution is to say that the T in the middle of the cluster is irrelevant, and that the S turns into shuh /ʃ/ to be more like the R that comes at the end: After all, both /ʃ/ and R are pronounced with the tip of the tongue pulled back somewhat. But yet another linguist, named Wayne Lawrence, argues that assimilation doesn’t work that way, skipping over some consonants in order to affect others.
The most likely possibility comes from still another linguist, named David Durian, who argues that instead of assimilating to a sound that comes later, the S is assimilating to a sound that comes earlier. In his study, he found that the words that were pronounced with a retracted S most often were those where the STR cluster came after the vowels /i/ or /I/, in words like destroy (deSHtroy) and restrict (reSHtrict). This is a very reasonable kind of assimilation, because these vowels are pronounced with the tongue raised high inside the mouth, and so is /ʃ/—at least, higher than it is for plain old S.
Why Do Pronunciation Changes Spread?
But wait a minute! I hear you saying. What about all those words that begin with STR? This gets into the second question, of why sound changes spread. Janda and Joseph argue that although a sound change begins with some kind of phonetic condition, if it hangs around long enough it begins to be affected by other things.
In this case, speakers seem to have generalized it from STR after certain vowels to STR anywhere—sometimes even across words. In fact, I even heard Michelle Obama, who does a lot of S-retraction, do it in the phrase parents trying, which she pronounced as parent[ʃ] trying. Janda and Joseph note that for some speakers, S-backing shows signs of generalizing even further, to any consonant cluster beginning with S. Examples they’ve heard include [ʃ]till, [ʃ]creen, and under[ʃ]tand.
Pronunciation Has a Social Aspect
In addition to spreading to different phonetic contexts, there’s also a social aspect to how sound changes spread. Durian’s study in Columbus, Ohio found that it correlated with a speaker’s pride in living in or having lived in an urban area, as well as with age: Younger speakers do more S-backing than older ones.
One more linguist, Kathryn Campbell-Kibler, looked at the social aspect from the opposite direction: How do others perceive speakers who have S-backing? Campbell-Kibler found that varying combinations of S-backing, S-fronting (otherwise known as lisping), and so-called G-dropping (for example, fishin’ instead of fishing) could affect whether a speaker was perceived as intelligent or unintelligent, gay or straight, from the American South, or from the country. One commenter on a blog post that I wrote on this subject said that S-backers were perceived as having lower social class in the UK; but here in the United States, another commenter reported hearing it from NPR reporters and on The Real Housewives of Atlanta.
So whatever generalization you’re tempted to make about speakers that use S-backing, don’t take it too seriously, because it’s probably more complicated than you think. And now that you know about S-backing, get ready to hear a lot more of it in your lifetime, as it gains [ʃ]trength and makes [ʃ]trides.