The main purpose of this demo is to illustrate the fact that in running speech the position of
the velum is not simply restricted to raised (for oral sounds) or lowered (for nasal sounds).
Rather, many intermediate positions may occur.
The film consists mainly of simple sound sequences at different speech rates, followed by
a series of real German words.
The film was made by inserting a flexible endoscope into the nasal cavity, allowing the
upper surface of the soft palate to be viewed. This is the bright moveable structure in the
lower part of the picture. The reddish area straight ahead is the rear pharyngeal wall.
To assist orientation in the film the word being spoken shows up as a subtitle.
In addition, the counter at the bottom of the picture shows the
running number of the speech item on the left (these numbers are used in the list below),
and a second/millisecond counter on the right that runs (roughly) while the item is being
spoken.
Credits: Endoscopy performed by Dr. Heidi Schröter-Morasch, Neuropsychologische
Abteilung, Krankenhaus München-Bogenhausen
List of the speech items, with comments
00 "swallow"
01 sustained i:
02 sustained a:
03 sustained sh
04 sustained sh, with crescendo
In these sustained sounds, all of them clearly not
nasalized, some differences in velar height can already
be noted (sh>i>a). Differences of this kind will become
even more apparent in some of the later sequences.
05 imi, ini, ingi (slow tempo)
06 ama, ana, anga (slow tempo)
Note that in these sequences with nasals, the velum
does not actually lower to its fully open position in the
nasals. Again, velar height is lower for /a/ than for /i/, so
there is less movement in the sequences with /a/.
A curiosity: In anga, at around 3s 900ms on the timer,
the effect of the tongue hitting the underside of the
velum when articulating the velar nasal can briefly be
seen. (Why isn’t this seen in ingi?).
07 imi, ini, ingi (fast tempo)
08 ama, ana, anga (fast tempo)
At the fast tempo, there is less movement than for the
slow sequences (especially for /a/), with the velum
staying in an intermediate position.
09 mamamamama (slow tempo)
10 mimimimimi (slow tempo)
11 mamamamama (fast tempo)
12 mimimimimi (fast tempo)
The remarks already made apply to these sequences,
too; i.e differences between /i/ and /a/, as well as
between slow and fast speech rate (note that
movement virtually disappears for mamama... at the
fast rate).
13 tanatanatana (slow)
14 tantantan (slow)
15 schmaschmaschma (slow)
16 tanatanatana (fast)
17 tantantan (fast)
18 schmaschmaschma (fast)
These sequences differ from the previous ones in
including a voiceless plosive or fricative. More vigorous
movement is now observed. Moreover, it does not
reduce so much at the fast speech rate. Thus it seems
to be much more crucial for “strong” consonants like /t/
or /sh/ to achieve a high velum position, than for a low
vowel like /a/.
19 Tinte
20 Tante
21 Onkel
22 Hamburg
23 fremd
In “Onkel” and “Hamburg” the velum raises only slightly
before lowering for the nasal, and then raising strongly
for the plosive, whereas in “fremd” there is strong
raising right at the beginning for /f/.
24 Mondnacht
This word “sandwiches” the plosive /t/ between two
/n/’s, showing how fast the velum can potentially move.
25 Eins
Similar to 21 and 22: very little velar raising before the
nasal
26 fuenf
27 Blinker
28 Lampe
Velar raising for /l/ at start of “Lampe” is slightly less
than for the initial /f/ in fuenf and /b/ in Blinker.
29 Hemd
Similar pattern to “Onkel” and “Hamburg” (21/22)
30 singen
Only very slight velar raising for the vowel between the
two nasals.