The Psychological Foundations of the Grammar 75
and the level tone to be closer than the semantic distance between either
the level tone and the fall, or the fall and the rise. However, contrary to his
hypothesis, his subjects considered that the level tone was as semantically
distant from the rise as it was from the fall, and that the semantic distance
between the fall and the rise was as great as the semantic distance between
the level tone and the rise. These fi ndings suggest that Gussenhoven’s
subjects treated level tone as a separate tone and not as a stylized variant
of the rise.
Tench (1997 and 2003) argues that a further and perhaps recent
communicative value realized by level tone is that of routine listing and
states that:
The pattern is often used in arguments when the speaker wants to give
the impression that they expect any self-respecting interlocutor to fully
agree with their statement without raising any objection. (2003: 229)He (1997: 17) provides the example of a doctor from East Anglia who, on
the Radio 4 Today news programme, said:
(36) some of the children are so \/ILL // that they can’t go to
— SCHOOL // they can’t even get up and — WALK //...
The doctor presents as self-evident the information that the children can’t go
to school, that they can’t even get up and walk and uses it to substantiate his argu-
ment. According to Tench, such instances of level tone have the potential
to operate as part of used language; speakers in pursuit of their individual
communicative goals package their message as a non-controversial routine
or list which they expect their hearers to agree with.
Before attempting to suggest how level tone may be encoded into a
grammar of used language it is worth summing up the above fi ndings.
There is widespread agreement that level tone labels utterances as routine,
detached from the context, and downplays the speaker’s involvement with
the message. Pickering (2001: 238) employs the term tonal composition to
refer to the combination of rising, falling and level tones in any discourse.
According to Brazil (1997: 135) a combination of predominantly end-falling
and end-rising tones labels the discourse as direct while a combination of
end-falling and level tones labels the discourse as oblique. (37) fulfi ls Brazil’s
intonational and grammatical criteria: the speaker produced a falling tone
and his words completed a grammatical chain but he is not apparently pro-
ducing used language and so (37) is considered an oblique increment.