The Ranums' Panat Times
Some further thoughts about (and images of) the vowels in French tonguing syllables
See also my Musing on the syllables TU and RU ( it is about wind syllables such as Veddy, Tu and Ru), for which this Musing is a companion piece....
I read with great interest Dan Laurin's article, "Shaping the Sound," in American Recorder, Sept. 1999, pp. 13-17also available online through Nicolas S. Lander's site. Laurin investigates "the color of one's tone" on the recorder and relates that color to "the shape of the vocal tract." I thank him for sharing his research with us; and I especially thank him for stimulating me to re-examine French wind articulations and the vowels on which they are based.
In my article on "veddy" versus tu and ru, I argue that a "pointed" French pronunciation of the wind articulations tu and ru affects a players tone by shaping the mouth in a specifically French way as contrasted with the Germanic-Anglo "too" and "roo" that is so often presented as an appropriate French tonguing. Owing to problems with Laurin's vocabulary (see my comment at the end of this article), I can't relate his article to mine, as I had initially hoped to do. Instead, I will provide some images that should not only clarify some of the points in my "veddy" article but that should also permit us all to delve deeper into the phenomena that Laurin is describing.
One of Laurin's main points is that by "creating more space in [his] mouth" as if "mimicking a giant yawn while playing" he could produce what he calls a "thin" sound. (This "thinness," he says, has nothing to do with the space in his mouth but to the quality of the sound. That is, the "noise" of the broad-band signal was reduced considerably, some of the partials were much more prominent, and the overall effect of this "thin" sound was quite "edgy.")
On the other hand, by creating "relatively less space in the oral cavity" as if "mimicking the letter 'a' in the word 'cat' " he could create a "thick" sound whose "sound spectrum," whose "timbre," is differ visibly and qualitatively from his "thin" sound.
Two differently shaped oral cavities that produce two distinctly different tone-colors, timbres, on a recorder played by a Swede who seems to be very much at home in English and who contrasts the results of mimicking a yawn and the word "cat."
And two differently shaped oral cavities in French wind articulations of the Baroque period, one created by the "pointed" u (shown throughout this article by the international phonetic symbol "[y] "and the other by "ooo," which is written ou in French and is indicated here by the phonetic symbol "[u] ". In sum, it is this choice between tu [ty] and tou [tu] and "veddy" that I described in my earlier article.
First of all, let's look at the ou [u], "ooo," that is used by so many of us.
What is the shape of the vocal tract when a recorder player mimics the French vowel ou [u]? Some diagrams in my friend Henri Morier's dictionary of rhetoric and poetics (Dictionnaire de poétique et de rhétorique, Paris: PUF, 1989 edition, p. 1281) show this shape from several angles.
If we start with the left image (a side view of the mouth), we see that just above the epiglottis, the throat measures 15 millimeters. (The epiglottis is that little upward-pointing flap that rises or lowers slightly for the different vowels.) The tongue rises quite high here, creating a 4.5 mm- high "channel" (chenal) just beneath the soft palate (the back part of the roof of the mouth). The sides of tongue push against the back teeth (darkened in the palatogramme) but the rest of the tongue is free. When viewed in cross-section (coupe médiane), the "posteriorized orifice" (orifice postériorisé, alas in very small print and hard to read...) that is created by the tongue proves to be quite small and flat.
NOTE: The 4.5 mm height of this channel, this orifice, contrasts markedly with both the 27 mm of the long French a [a] /"ahh" and the 18 mm channel of the shorter French a [a]. The latter of these two a's comes pretty close to the a in Laurin's "cat." (Morier shows these vowels on pp.1274-75, but I'm not reproducing those illustrations here because they appear in a simpler form in my fourth illustration).
It is through this orifice, located rather far back in the mouth, that the air moves from the throat into the front of the mouth and then into the recorder.
Since the rear of the tongue is very high, the tip of the tongue does its articulating rather far back in the mouth, at the dotted line marked "22 mm." Depending on the consonant with which the speaker launches his ou [u], the tip of the tongue has to move up or down approximately 22 mm if it wants to alternate between syllables that articulated near the roof of the mouth, and syllables that are articulated near the lower teeth. In other words, to articulate tou [tu] the tongue has to touch the flesh below the lower teeth; but it articulates a rou [ru], "roo" or a dou [u], "doo" on the front ridges of the roof of the mouth.
Now let's look at the pointed French u [y] .
Sounding this "pointed" u creates a very different vocal tract (p. 1276). First of all, just above the epiglottis the throat is somewhat more open than it is for ou [u]: 17 mm compared to 15 mm.
Second, the tongue arches toward the roof of the mouth (palate) considerably farther to the front than for ou [u]. This enlarges the space in the rear of the mouth, but it greatly restricts the space in the front of the mouth which varies from 5 mm high at its narrowest point to 6 mm just where the tip of the tongue touches the front teeth.
In order to lift the tongue so high, the sides of the tongue virtually glue themselves to the upper teeth, as indicated by the darkened regions of the palatogramme. The result is a very narrow and flattened "anteriorized orifice" (orifice antériorisé... again the words so small that they are hard to read) located rather far forward in the mouth. Through this orifice the air from the lungs passes in order to exit from the mouth and enter the recorder.
Of special interest here is the position of the tip of the tongue. It rests on the lower front teeth, not on the flesh below the gum as it does for ou [u].
Finally, let's look also at the French "eh" that corresponds more or less to the "veddy" of which I am so critical.
As the illustration shows (p. 1273), this e narrows the throat above the epiglottis, flattens the tongue and leaves a channel between the tongue and the roof of the mouth that is considerably larger than the orifices of u [y] and ou [u]. In fact, this open e creates the largest of the three openings shown in this illustration. This opening is 9 mm high, compared to the 4.5 and 5 mm heights of [y] and [u]. The tongue touches the back teeth (the darkened area on the palatogramme) just a wee bit more than for the French ou [u]. But any up-and-down motion of the tongue involves only 12 mm, rather than 22 mm.
Even more conveniently for any player who wants to alternate "ved-dy, ved-dy," the tongue rests at the top of the lower front teeth, permitting a relaxed alternation between "ved" (just at the top of the lower teeth) and "dy" (at the first palatal ridge just behind the upper front teeth).
To summarize and contrast:
When a windplayer uses the "pointed" French u [y], he relaxes the back of his throat; but he pushes his tongue upwards into such a tight position that the "orifice" through which air exits from his lungs is not only quite far forward in the mouth, it is extremely narrow. He is forced to articulate his syllables in the very front of his mouth, fluttering his tongue back and forth a mere 6 mm as he alternates tu and ru. This is a very efficient articulation, one that lends itself to rapid passages as well as slow ones. The concentrated, focused flow of air that exits the mouth into the recorder surely affects the tone, the color, the timbre of the note, a la Laurin?
By contrast, if a player says "too, roo, doo," the position of the tongue causes him to narrow the back part of his throat and open the front of his mouth. This creates a somewhat wider "orifice" through which the air can enter the mouth; but the orifice is farther back in the mouth. (So I suspect that, if measured in a lab, the wind flow would be somewhat less intense than for a [y]?) Now, if a player tries to alternate tou and rou ("too" and "roo") his tongue has to cover 22 mm between each syllable! This tonguing is so slow and so inefficient that most players would rule it out for rapid passages. This, I suspect, explains why French wind tutors do not propose "tou" and "rou" but prefer "tu" and "ru."
The clumsiness and the slowness of tou and rou doubtlessly explains why someone (circa the 1960s or 1970s?) concluded that the French Baroque r was really more like a d than an r. (I have found no historical evidence that would support this premise, but there is ample evidence in the sources to contradict it.) Exiling the r conveniently eliminated the upward tongue-curl that is required to form the French seventeenth- and eighteenth-century lingual r. And it reduces the distance the tongue must travel between different syllables. That is to say, since t and d are articulated in almost the same place (just behind the upper front teeth), "too" and "doo" require minimal up-and-down motion. This way of avoiding the "pointed" u is therefore quite efficient. But it presents the distinct inconvenience of not even remotely resembling the tonguings discussed in French handbooks.
There is another still more efficient tonguing: "ved-dy." In this tonguing pattern, the ou is replaced by alternate "eh"s or an "eee"s, so that the tongue needs move only 12 mm up and down, just behind the front teeth. This is a convenient, efficient invention; but it has nothing to do with the French articulations described by the sources. The "orifice" is much too large, the consonants are wrong, and the back of the tongue is much too relaxed.
What are we to conclude?
First off, it would be wise to conclude that "veddy," with it's e, should be eliminated as a plausible simulacre of French articulation to be used by non-French speakers.
That leaves the "pointed" u [y] and the broader ou [u]. The latter is so fraught with problems that it scarcely merits further attention. The pointed u is therefore the only viable vowel to imitate.
French articulations and Laurin's article...
Are tu and tou differ enough from one another to create the sort of contrasting sound spectrums that Mr. Laurin describes?
He provides so little information about the shape of his mouth and throat during his experiment that it is difficult to judge. We do know that he thought "cat" for his "thick" sound, so this sound probably was created with a vocal tract similar to the one marked "a" in the upper-right corner of the following chart from Morier. Still, Laurin describes this "relaxed" position as creating "relatively less space in the oral cavity" than his "thin" sound did. The latter, he says, is produced by imitating a "great yawn," thereby "creating more space in [the] mouth." Since the two a's in Morier's chart involve the largest oral cavities of all French vowels (and since it seems unlikely to me that English is spoken with markedly larger cavities than French), it is difficult to imagine how Laurin could articulate a consonant if he enlarged his oral cavity considerably more than one would for the a of "cat."
Come to think of it, Laurin does not discuss consonants! Did his experiment involve measuring un-articulated sounds? Did it divorce the two components of a wind syllable: the consonant that tongues the note, and the vowel that gives the note its length and its timbre? Did he measure a "great yawn" without giving consideration to which consonants can be articulated in that position, and which one's can't? For example, I believe that I have managed to reproduce something close to his "great yawn" on C6; but I can't manage to launch that yawn with an r; nor is my attempt at a t very felicitous. And I appear to be incapable of subdividing my "great yawn" into quicker notes by alternating r and t or even t and d: the sound wavers at each attempted articulation, but no separate syllables can be discerned.
In fact, my attempts at articulating a "great yawn" call to mind the singers whom I coach in French Baroque rhetoric. Their vocal teachers have taught them to focus on vowels, and above all to convert those nasty French vowels into rounder, more melodious sounds. And they have been encouraged to keep their consonants to a minimum, lest this disagreeable sounds impinge upon the beauty of the vowels. To make these singers intelligible, I have to turn that approach completely around ! I have to encourage them to linger on the consonants and let the vowels take care of themselves.
But let's return to Dan Laurin's article. To understand fully the phenomena about which he is writing, recorder players will have to await his projected "sound method," which doubtlessly will clarify these issues. Meanwhile, in the interests of scholarship I have taken the liberty of borrowing the following chart from a publication by my dear friend, Henri Morier (Dictionnaire de poétique et de rhétorique, Paris: PUF, 1989 edition, p. 1258). It shows the vocal cavities for the different French vowels:
To what extent can Laurin's findings be carried over to French Baroque performance practices? At this stage in his presentation, it's not clear. Still, I should think that the sound spectrum and the force of the air exiting from the mouth and entering the recorder would be different for tu and for tou / "too"? I should think that mimicking these two vowels would produce two distinct tonal colors?
Until Laurin and other scientifically minded musicians tell us more, I'll close by raising an esthetic point. Does not Laurin's experiment reveal something about changed musical esthetics over the past few hundred years? And from nation to nation?
In order to get a "nice and musical recorder sound," he mimicked "cat." As far as I can judge, an articulation syllable based on either of the French a's diagrammed above would have astonished 17th- and 18th century French musicians. "But, Monsieur Laurin, saying ta or ra" leaves the mouth wide open!" they would doubtlessly have exclaimed. "How are we supposed to articulate the notes? That's why we say tu and ru: it's to keep the lips pursed ..."
Two very different musical esthetics appear to be involved in my research and Laurin's. The instrumental sound that I am describing relies on the tense, "pointed" mouth so characteristic of the French language, which tends to focus its sounds in the front of the mouth. Laurin's ideal sound which involves imitating an a with a "relaxed" mouth belongs to a very different linguistic world. Though he writes about reducing the size of the vocal tract, both of the vowel-like positions that interest him (the "nice" and "musical" a and the "great yawn") are located low in the mouth, far back in the throat. Indeed, the contrast between the two esthetics calls to mind the contrast between the two singing styles that I evoked above. The music of seventeenth-century France clearly was not guided by the same parameters as the multi-cultural and multi-lingual world of twentieth- and twenty-first-century musicians!
Some clarification about how consonants affect his choice of vowels (see above) would also be helpful.