Chapter 7. Ejectives in English and German 197
account. In particular, Kingston (1985) shows using a digital implementation of
Rothenberg’s aerodynamic model (Rothenberg 1968; Müller & Brown 1980) that
larynx raising alone is not able to produce the necessary increase in supraglottal
pressure to produce the types of ejective bursts he observes in languages such as
Tigrinya. Instead, he suggests that additional articulators may be employed to
reduce supraglottal cavity size, such as tongue root backing.
However, in an attempt to find an adequate account of the production mecha-
nism for epiphenomenal ejectives in German and as a consequence also ejectives
in English, we must even cast doubt on the initiatory contribution of decreasing
the size of the supraglottal cavity to increase intraoral pressure and fuel an ejective
plosive release.
In German, epiphenomenal ejectives are the product of a temporal overlap of
a final plosive and junctural glottalisation, i.e. glottal stop or creak at vowel onset.
Figure 4 compares (a) an epiphenomenal ejective release from weht ein ‘blows a’
with (b) the pulmonically fuelled release of a final pre-pausal plosive in mit ‘ w i t h’
from Simpson (2007). This paper looks at different ways in which both nasal and
oral stops in German can be produced with releases fuelled by non-pulmonic
airstream mechanisms. So, for instance, in nasal–plosive sequences such as that in
in Kiel, we routinely find a weak, yet consistently present, click at the point of the
apical release of the nasal. This click is an epiphenomenon produced following a
double, apical-velar closure followed by the release of the frontmost (apical) clo-
sure. A sufficient change in air pressure needed to give rise to the click is produced
by a small change in the size of the intraoral cavity prior to apical release (Ohala
1995). Likewise, in line with Ohala (1997), Simpson (2007) suggests that in the
ejective release of plosives we find in an example such as that shown in Figure 4a,
a change in supraglottal air pressure is brought about by vowel-to-vowel move-
ment taking place during the double glottal and oral closure. However, there are
two things that might lead us to question this interpretation. First, pressure change
due to vowel-to-vowel movement predicts that we should find plosive releases
fuelled by both glottalic egressive (ejectives) as well as ingressive (voiceless implo-
sives). In the first case, intraoral pressure increases due to a vowel-to-vowel move-
ment bringing about a reduction in the size of the supraglottal cavity, e.g. [a–i].
In the second case, intraoral pressure decreases due to an enlarging vowel–vowel
movement, such as [i–a]. However, no voiceless implosive releases were found.
Secondly, impressionistically, many such plosive releases seem to be intenser than
might be expected from a lot of vowel-to-vowel movements, e.g. [e] to [a] in weht
ein. Although these doubts are based primarily on auditory impression and visual
interpretation of the acoustic record, it does seem worth speculating about other
production mechanisms behind such ejectives.