One of the reasons I don’t post here as often as I’d like — certainly not as often as I think about it — is because I spend a lot of time answering questions for people. This morning, I received and answered an email from a wonderful tutor I trained last year in Pennsylvania, a deep thinker whose approaches her language scholarship with integrity. I liked her questions — and my answers — enough to massage it into a post. One of the biggest problems in phonics and in literacy research is a flawed understanding of phonemes. This tutor’s email betrays a sophisticated understanding not only of what phonemes are, but also of the difference between phonology and phonetics, something that phonics is hard-pressed to explain.
She began her email with a question about diphthongs:
“Vowel phonemes that are comprised of 2 phones are called diphthongs, like the /aɪ/ in wide or the /ɔɪ/ in boy.”
She continued: “Do most long vowels slide?”
Indeed, I responded. English has 6 ‘long vowels':
/eɪ/ /iː/ /aɪ/ /oʊ/ /ju/ and /uː/, as in bait, beet, bite, boat, beaut, boot .
Only /iː/ and /uː/ do not glide. The rest of them do. You may note that there are kind of two “spots” in a long vowel phoneme. When we write /iː/ it’s like we were writing /ii/ — where the phone [i] is taking both the “spots” in the long vowel phoneme. The gliding vowels have different phones in each of their “spots.”
But “I don’t hear it [the glide] in long <o> or <oo> or <e>, ” she confessed.
She’s aboslutely correct that there’s no glide in ‘long oo’ or ‘long e’ — that’s /uː/ and /iː/ in the International Phonetic Alphabet. I explained that the glide in /oʊ/ [‘long o’] is slight, but it is there. It’s not that we don’t hear it; it’s that we don’t perceive it. British English glides even more noticeably in this phoneme — I’d represent British ‘long o’ as /əʊ/, American as /oʊ/. If you try to pronounce “go” the way you normally would, and compare it to someone saying “go” with a French accent, you may be able to then perceive the glide in the English version.
She went on. “Would you teach students for spelling the concept of sliding vowels?”
So, I did not say this in my email but I want to highlight here that this is a question I frequently get: “Would you teach students X?” All I can really say is that I would share with students any understanding I have about language, bar none. At no point does it make sense to me to deliberately withhold facts about the writing system from anyone. Now, the question is a charitable one, and I understand the time constraints and lesson planning decisions that drive it. But no one will ever hear me say, “No. That is a piece of understanding language I don’t recommend sharing with children.”
Okay, back to the question. I would absolutely teach children about diphthongs. Goodness knows they’re already mistaught plenty about diphthongs. Children frequently misspell words precisely because they’re perceiving a diphthong, such as when they write cycle as <saycl> or white as <wait>.
The tutor switched gears. “Consonant phonemes that are comprised of 2 phones are called clusters, like the /ʧ/ in chip or the /ks/ in box.”
Right again. And, while I didn’t say it in the email, I’d like to add that this is something that phonics gets dead wrong, pretty much every time. They’ll tell you that <x> spells two phonemes — /k/ and /s/. It’s because they don’t understand the crucial difference between phones and phonemes: <x> spells a single phoneme that is comprised of two phones, a consonant cluster phoneme. Many languages have consonant cluster phonemes, like /ʦ/ or /ps/: the /ps/ phoneme in Greek is spelled with a single character, <ψ> (psi), just like our <x>. If something is spelled with a single grapheme in an alphabetic writing system, it’s a single phoneme.
But “how does /ʧ/ have 2 phones?” she asked. This is why I love her question: because she’s not just trusting the definition I give her, or accepting that the /ʧ/ is two phones because I said so. She’s looking for an understanding she can call her own, so she can do just what she asked if she should do: share this understanding with her students.
Here’s what I told her: [ʧ] starts like a [t] and ends like a [ʃ]. Think of it this way — say the word “hatchet” and feel the /ʧ/ in the middle. Then, say the word “batshit” — as in “she’s batshit crazy” — and feel the /tʃ/ in there. You don’t have a hard time perceiving both the [t] and the [ʃ] in “batshit” because it’s a compound and you’re aware of the morphemic boundary between the two phones. But because you perceive (as a native speaker) the /ʧ/ in “hatchet” as a phoneme — as a unit — it’s harder for you to be aware of the two phones. The [t] is unreleased — if you pronounce “bat” without the “shit” — you probably release the /t/ at the end, so it is aspirated as [tʰ]. But when you pronounce the /ʃ/ right next to it, it’s not released, so it sounds like a /ʧ/. This should help you be able to better perceive the two phones in /ʧ/.
Now, I’m not suggesting that hatchet and batshit are phonetically identical. It’s a difference between /ʧ/ and /tʃ/. It may be tacky, but this batshit/hatchet pair provides an excellent example of how /ʧ/ has two phones, because they rhyme. A kid-friendly example might be ouch and outshine, or studying the /tʃ/ in hotshot, courtship, or nightshirt. Understanding the structure of the /ʧ/ also helps us understand why some kids write words like tree and trap as *<chree> and *<chrap>: the /t/ in trap is phonetically closer to the [ʧ] in chap than it is to the [tʰ] in tap. It’s the same phenomenon when they write *<jress> instead of <dress>.
More importantly, studying this pair also peels back the ways that orthographic understanding runs hand in hand with true phonemic understanding. Being aware of phonemes is all well and good, but awareness without understanding is a little crazy.