Posted by Neal on September 19, 2008
Last month, I said
I’ll have more to say about “r-colored vowels” … on September 19.
Now that it’s September 19, it’s time for me to do that. As I wrote in that earlier post, I don’t like how phonics materials, and even phonetics texts, represent the er in landlubber or the ur in scurvy. Sometimes they call them a schwa or short U plus /r/, and sometimes they use the combination symbols [ɚ] (for the unstressed syllable) and [ɝ] (for the stressed one).* To me, though, this syllable just sounds the same as the [r] sound that you get in words like rum, except that instead of being used as a consonant, it’s being used as a vowel. Or in phonetic terms, it’s being used as the nucleus of a syllable.
When a consonant is used this way, it’s referred to as a syllabic consonant. Syllabic consonants are represented in the IPA by a small vertical line under the character, but since I don’t know how to do that in Unicode, I’ll use an underline. Here are a few words with syllabic consonants in English:
- [brn] burn
- [mɪʔn] mitten
- [baɾl] bottle
- [phst] psst!
- [s̃ ʌʔm] something (pronounced sumpm)
There are two consonants that have their own designated vowel character for their syllabic versions. Instead of saying “syllabic W”, we just call it a long U ([u] in the IPA). Instead of saying “syllabic Y”, we call it a long E ([i] in the IPA).
Fine, you say. Maybe the vowels represented by er and ur (and for that matter ir as in mirth and ear as in pearl), should all just be considered syllabic [r] in American English. What about other vowel+R combinations, like the ones in Arrr!, corsair, ashore, tour, and buccaneer? Antony Dubach Green has an interesting paper on this very subject. Before I give his claim, some background is in order. I’ve talked about diphthongs before. The most commonly discussed diphthongs in English are [ai] (“long I”) as in Aye, [au] as in now, and [ɔi] as in Ahoy!. Even materials for nonlinguists recognize these as diphthongs. Less easy for English speakers to recognize are the “long A” and “long O” sounds, which are usually pronounced with an [i] or [u] at the end. So hey would be transcribed [hei], and yo ho ho would be transcribed as [youhouhou].** So far, all these diphthongs end in [i] (a high front vowel), and [u] (a high back vowel). Seldom mentioned at all in the introductory phonetics texts I’ve seen is the fact that some diphthongs end in neither a front nor a back vowel, but a central one, specifically [ə]. This typically happens before [l], as in keel, which would be transcribed [khiəl̴]. (Also in Neal). This kind of diphthong is known as a centering diphthong.
So at this point, two vowels that also serve as consonants appear as the ending vowel in English diphthongs: [i] and [u]. Why don’t we also have diphthongs that end in our other vowel that also serves as a consonant, namely [r]? That, in fact, is the gist of Green’s claim, which builds on a similar claim from John Harris’s 1994 book English Sound Structure. He concludes that some Vowel+[r] sequences are actually diphthongs that end in [r]. Specifically, they are the r-colored vowels in Arrr!, corsair, and ashore. He also argues that the r-colored vowel in scurvy and landlubber is a diphthong, but that’s where he and I differ, since (as I said earlier) I’d call it a syllabic [r]. However, Green believes that sequences of a high vowel followed by [r] — in other words, the [ir] and [ur] sequences in buccaneer and tour — really do consist of a vowel followed by [r], in the same way as you can have two vowels in a row that don’t form a diphthong, as in rio.
What is Green’s evidence for this claim? To support the claim that at least one r-colored vowel behaves like a diphthong, he cites a 1990 study by A. C. Cohn, who observed that after a speaker pronounces a nasal consonant ([m, n, ŋ]), the nasal passage closes back up again over the course of one vowel. Diphthongs act as a single vowel, so that in a word like night, the [i] part of the [ai] diphthong is still nasal; that is, air is still exiting through the nose. Cohn noted that the [r] in more was also still nasalized, indicating that the [ɔr] sequence is acting like a diphthong. However, the [o] in neo and the [r] in near are not nasalized, indicating that [io] and [ir] are not diphthongs, but two-segment sequences.
As for the rest of the r-colored vowels, Green offers evidence from a language game called Uzzlefuzz, in which every syllable of a word is replaced by V+uzzle+V, where V represents the original vowel in the syllable. Thus, keel would be keezlefeel, and booty would become boozlefooteezlefee. Diphthongs count as a single vowel, so aye would be [aizlfai], not [azlfai]. The words nurse, sharp, scarce, and north treat the r-colored vowel like a diphthong, resulting in nurzlefurse, sharzlefarp, scarzlefarce, norzleforth. But [ir] and [ur] are treated as separate segments, so that fierce and poor (presumably pronounced to rhyme with tour) become feezlefierce and poozlefoor, and not fierzlefierce and poorzlefoor.
So avast, me hearties, and welcome aboard syllabic [r] and its r-final diphthongs [ar], [ɛr], and [ɔr]!
*Since I’m doing a whole post about [r], I should note that [r] is not the IPA symbol for the (American) English R sound. In the IPA, [r] represents the trilled R, as in Spanish perro, while the English R is represented as [ɹ]. However, since I won’t be talking about both of these sounds, I’m just not going to bother.
**The consonant Y sound is not represented as [y] in the IPA, but as [j]. For easier reading for a nonlinguist audience, I’m using [y] anyway.