36 Advances in spoken discourse analysis
expected’; ‘complacent, self-satisfied or smug’; ‘disclaiming responsibility,
shrugging aside any involvement or refusing to be embroiled’.
It soon becomes evident that some, perhaps much, of the claimed
attitudinal meaning is, in fact, being derived from the lexico-grammatical
and contextual features of the examples themselves and not from the
intonation contour.
Thus, although there is no disagreement that speakers can vary independently
tempo, loudness, pitch, and voice quality, and thereby alter aspects of the
meaning of their utterances, one must conclude that any systematic relationship
between intonation choices and lexical meanings has so far remained
undiscovered. Indeed, Labov and Fanshel imply that a search for systematic
relationships is misguided when they suggest that the lack of clarity or
discreteness in the intonational signals is not ‘an unfortunate limitation of
this channel, but an essential and important aspect of it’ (1977:46). The
result is that, in the absence of any satisfying theory to account systematically
for the interactional meaning of intonation, those involved in the analysis
of spoken interaction have, of necessity, taken only intermittent notice of
intonation choices, at those points where they felt they could attach significance
to them.
Perhaps the paradigm example of this approach to intonation is the way
in which Sinclair and Coulthard (1975) used the co-occurrence of the prosodic
features ‘high falling intonation’ and a ‘following silent stress’ with ‘now’,
‘well’, ‘OK’, ‘right’, ‘good’, to isolate occasions when these lexical items
were functioning as ‘frames’, markers of boundary points in the ongoing
lesson.
More generally, most analysts have felt able, as native speakers, to
recognize, though not necessarily to describe, the intonational features
that mark certain declarative clauses as questions and certain words as
‘stressed’. Indeed, Jefferson’s (1978) transcription system, which sets out
to be ‘one that will look to the eye how it sounds to the ear’ (p. xi), also
allows for a ‘continuing intonation’, and a ‘stopping fall’, plus three degrees
of stress. However, as none of the published transcriptions have an
accompanying tape and as only Labov and Fanshel provide fundamental
frequency traces, it is impossible to be sure what phonological features
particular analysts are focusing on, how consistently they are recognizing
and marking them, how much agreement there is between analysts on
what constitutes a question-marking intonation or a particular degree of
stress, and how far it is the phonological features alone to which they are
responding.
Thus, it is evident from the use made so far of intonational information
in published work that all those involved in the analysis of verbal interaction
would agree with Labov and Fanshel (ibid.:46), that it is at the moment
impossible ‘to provide a context-free set of interpretations of prosodic
cues’.