192 Adrian Simpson
Ogden (2009) describes and illustrates tokens of ejectives using examples
from naturally occurring talk from a Scottish English speaker. Ogden’s examples
illustrate well the different interactional functions ejectives might be fulfilling.
In common with other descriptions, ejectives occur finally in his data, but the
structural intricacy of conversational data allows for a more detailed analysis. So,
for instance, ejectives seem to be one of the correlates of floor-holding pauses,
first described in Local & Kelly (1986) in a data set initially described by Jefferson
(1983). Such a floor-holding pause was found to begin with glottal closure and
end with its release. The ejective shown in the example in (1), taken from Ogden
(2009) would seem to be part of such a pausal complex.(1) at– [[ʔ] on the week]end of wee[k’ (0.3 s) ʔ] three (Ogden 2009: 165)In (1) the closed glottis extends from the end of the word week to the beginning of
the word three. It would be possible to interpret the ejective release of the dorsal
plosive as an explicit expression of the glottal closure.
These two analyses of ejectives from Gordeeva & Scobbie (2006) and Ogden
(2009) highlight a possible conflict between two theoretically different interpre-
tations of the same data set. In Ogden’s approach, a range of different phonetic
patterns are accounted for in terms of the work they are doing in structuring
conversation, and variation per se plays a subordinate role. By contrast, while
Gordeeva & Scobbie’s analysis does take account of structure (e.g. finality), a
more detailed categorization in terms of interactional structure is not present –
the concentration is on ejectives as part of the possible set of patterns of varia-
tion. However, analysing sociophonetic variation using only restricted contextual
information, such as finality, is dangerous. Assigning different phonetic shapes
to the same sociophonetic variable assumes that information about different and
identical structural context is known. But as Example (1) illustrates, within nor-
mal interaction, there are different categories of word-finality or pre-pausality,
different contexts which may or may not be accompanied by different bundles of
phonetic events. Another well-documented example of this is the high prevalence
of word-final plosive aspiration found in Tyneside English speakers producing
word-list material (Docherty et al. 1997; Local 2003). A possible sociophonetic
interpretation of this, but one which only considers word- or utterance-finality as
a structural categorization, is that the higher frequency of occurrence of aspirated
release is an approximation to standard forms. However, in an earlier study on
the phonetic shape of turn-taking in Tyneside English, Local et al. (1986) found
that the same phonetic pattern, i.e. the aspirated release of voiceless plosives, is
one of the phonetic correlates of turn-finality. It is hardly surprising, therefore,
that speakers of this variety of English should produce aspirated final plosives in
word-lists, producing the phonetics of turn-finality after each word (Local 2003).