Didja Eat Yet? How About Cheese Pizza with Buttons and Tsetses?

note: the intelligibility of this post is riding on fonts. Here's hoping you can see the following symbols č ǯ ʃ ʤ ð ʌ and ʔ. If you cannot...I apologize. I'm not enough of an expert to fix it yet but I'm doing my research. (Though a good place to start is by using Mozilla/Firefox. It's a good idea anyways.) If you can read the symbols, all's set; the only problem now will be my horrible writing.

No system that claims to be all encompassing is without its detractors. The International Phonetic Alphabet is under siege from quite the linguistic nerd-horde. In one of my classes a very capable professor decided to swing his hatchet at the ankles of the IPA's notational convention for affricates. Here's the contention:

Affricates (such as the sounds that begin the words choose cheese chair june jack jeep) are transcribed in the Americanist phonetic system as a single sound [č] in cheese and [ǯ] in jeep.

They are transcribed in the IPA as two sounds with the symbol combinations [tʃ] and [ʤ]. Using standard orthography the sounds might be interpreted as "tsh" and "dzh."

We will focus today on the voiceless affricate: [č]/[tʃ].

The fine professor I mentioned above claims this is not an accurate notation. The affricates are not two sounds, he says, but one. His tongue in cheek explanation of the motivation to represent them as two is the influence of French language on phonetic notation. There are no affricates in French so they must not see the possibility of just one sound like that. It would be similar to the English lack of the alveolar affricate [ts]. This is the consonant sound we hear twice in the name of the fly whose bite causes sleeping sickness: the tsetse fly. In English it is natural to see this as two sounds.

His case for the "one sound" analysis: say the phrase "white shoes." According to him this results in the [tʃ] juxtaposition and clearly reveals two separate sounds. The phrase would be transcribed [waɪtʃuz] He then suggests a relevant phrase to illustrate the affricate as one sound: "why choose." "Can you hear the difference?" he asks. Heads nod. He nods once to confirm his own argument's validity. He transcribes it thus: [waɪčuz].

Here's where I see his argument falter. One difference between the phrases is the first vowel diphthong. In "white" there is a mid central vowel [ʌ] (as in cut or thus) followed by a high front vowel [ɪ] or [i]. In "why" the diphthong is a low vowel (either front [a] or back [ɑ] depending on analysis and dialect) followed by roughly the same high front vowel. Notice the difference in vowel sounds between "writer" and "rider." "Rider" is a lower and even quantitatively longer vowel. These diphthongs are the first difference the students hear.

Setup to problem two: the glottal stop is the stop of air we hear when we say "uh oh" "uh uh" or the sound we hear instead of a [t] in American "button." It's that popping grunting noise that we make by closing our throat. It is phonetically notated with an undotted question mark: [ʔ].

Thus problem two: our venerable professor tells the students to ignore any possible glottal stops as an influence on the pronunciation. But when everyone in the room says "white shoes" the glottal stop is definitely a part of the sound heard. In fact most students saying "white shoes" don't even use the [t] sound in the affricate: the glottal stop is followed directly by the post-alveolar fricative [ʃ]. I.e. the tip of the tongue never touches the little ridge behind the teeth. When saying "why choose" there is no glottal stop. This is why his argument falls apart. He is asking the class to compare what he claims is a [tʃ]/[č] difference but when they all hear and are actually comparing [ʔʃ] and [tʃ] he tells them it's because one of the affricates comprises two sounds and the other is a single sound. I see them both as two sounds.

One difference between the two sound combinations ([ʔʃ] and [tʃ]) is the pause or gemination that occurs after the glottal stop (first combination) and which does not occur after the alveolar stop (second combination). This is probably what leads him to see the first combination as more clearly two sounds. Instead of transcribing it simply as [ʔʃ] we may concede that in the phrase "white shoes" the consonant juncture is actually transcribed [ʔ:ʃ]. So yes - of course this combination at the word boundary is composed of two consonant sounds. And yes we can clearly hear them both because of the gemination that helps delineate them. But I stand firm that "choose" begins with two consonant sounds [tʃ].*

*Bonus boring stuff: For what it's worth say that last sentence out loud. Focus on "that choose." When it is read conversationally I would transcribe it this way [ðəʔ:tʃuz]. Right now I don't know if this helps my case or muddies it.