56 A Grammar of Spoken English Discourse
a particular feature on B’s map. He provides example (5) as support
(ibid. 38).
B B/A B/A/B
(5) A: you start below the Palm Beach, right + +
B: right + + +
A: you go over to quite a bit belowThe ‘+’ symbol in the fi rst column on the left indicates that B, who is
looking at his/her own outdated map, knows that the Palm Beach is on the
map. In the second column it indicates that B believes that A believes that
the Palm Beach is on B’s map. The evidence for this belief is the fact that A
has referred to the Palm Beach as being on B’s map. In the third column it
indicates that B believes that that A believes that B believes that the palm
tree is on B’s map. B has provided evidence for this by confi rming that A
was correct to believe that the Palm Beach was on B’s map.
This view closely resembles the truncation heuristics proposed by Bach
and Harnish, but has the advantage of recognizing that communication is
an inherently risky undertaking. Speakers and hearers often operate at
cross purposes with no guarantee that they form correct assumptions about
the extent of their shared beliefs. Clark and Marshall’s criticism that limited
regressive checks cannot guarantee mutuality does not apply because Lee,
unlike Bach and Harnish, does not argue that mutuality is a prerequisite for
successful communication.
Problems, however, still remain. The fi rst relates to the data, Brown’s map
task which, although described by Lee (ibid. 21) as authentic, is hardly
representative of most communication (see also Halliday and Matthiessen’s
(1999: 328) point that a considerable amount of language is not instrumental).
It is not at all clear how experimental designs such as the map-task could
be adapted to measure shared beliefs when speakers produce evaluative
language. Halliday and Matthiessen (2004: 34) caution that what people
say or understand under experimental conditions is very different from
what they say or understand in real life. The second problem is that an
individual’s belief is not a physical thing^8 and it is unclear how an individual
can share another’s belief. The third problem is that Lee defi nes belief
as less than 100% certainty in the truth of a fact but does not suggest
a threshold below which the speaker no longer believes. Nor does he pro-
pose a mechanism detailing how speakers are able to calculate the strength
of hearers’ beliefs.