Real tongue constrictions of finite extent alter resonances differently from idealized thin walls.

Evidence

In the vocal tract, constrictions cannot take the form of thin rigid membranes, with the possible exception of the lips and teeth. The back of the tongue is quite flexible, but any constriction it makes will have some extent along the tract. Such constrictions, with finite extent along the tube, impose some modifications to our conclusions based on the thin constrictions: they reduce the length of chambers residing on either side of the constriction. A constriction at the antinode may now rise in frequency, whereas it held steady before for the case of the thin wall. In addition, the constriction of finite extent at the node will still cause the frequency to fall, but not quite so far as for a thin wall. — Eric J. Heller, Why You Hear What You Hear- An…, 1645

Categorization

• QUESTION: Does a finite-extent constriction at a node shift formants in ways the vocal-tract model must always correct for, or does the correction stay negligible in practice?
• DEFINE: Clarifies the distinction between idealized thin constrictions and physically realized finite constrictions in vocal-tract acoustics.