200 Adrian Simpson
provide any evidence of ejective fricatives. The epiphenomenal account predicts
this. It implies a pulmonically driven intraoral pressure build-up prior to glottal
closure and finally oral release of stop closure. A similar sequence of glottal events
synchronised with the stricture of close approximation of a fricative will produce
pulmonically fuelled friction followed by its cessation once glottal closure is made.
Interestingly, final ejective fricatives would also appear to be absent in English, as
well (Ogden 2009).
A second piece of evidence suggesting a pulmonically fuelled pressure build-
up is the data from Suffolk English presented above. In Figure 2 the spectrogram
is aligned with the EGG signal. What one might have expected from the EGG
trace is evidence of differences between the two plosive types prior to stop clo-
sure. However, this was not the case. Glottal activity observed from the EGG sig-
nal showed no obvious differences for the ejective or the aspirated stop releases.
Again, if we assume that there is a pulmonically driven build-up of intraoral pres-
sure during the initial phase of the plosive, the only difference between the two
plosives would reside in whether the glottis was closed or open on release.
Although it is outside the remit of this paper, it is worth considering to what
extent languages with ejectives as part of their regular phonological systems may
also employ a similar mechanism. Indeed, in Rothenberg’s aerodynamic model
emploved by Kingston (1985) to model various aspects of ejective production
that was mentioned earlier, one set of model configurations employs a constant
glottal flow to raise sufficient pressure for the ejective burst, i.e. it does not rely on
the reduction in the size of the supraglottal cavity alone to produce the required
change in air pressure. Furthermore, the presence of perseverative voicing well
into the stop closure phase of all plosive types including ejectives in Georgian
(Wysocki 2004; Grawunder et al. 2010) implies that at least part of the pressure
build-up in some ejectives is pulmonically fuelled.- Discussion
Unlike epiphenomenal ejectives in German, ejectives are still an elusive feature
in many varieties of English and for many speakers, as is evidenced by their rela-
tive paucity in the data set from the first series of The Office. Despite being able
to coarsely describe some of the linguistic and interactional contexts in which
they occur, predicting their occurrence for a particular speaker in a particular
context is far from straightforward, and it remains a challenge to see whether
it will be possible to provide a chronologically ordered sociophonetic analysis
of the regional spread of ejectives through English. At present this is a daunting
task. Despite there being several acoustic databases of English covering different