Thursday 29 April 2021

Sensing vs Saying: Constituency Boundaries

Halliday & Matthiessen (1999: 130-1, 131n):
Figures of saying construe the addressee of exterior symbolic processing in the form of a participant, the Receiver, as in She told/asked/commanded him…; She said to him/ asked of him… . In contrast, interior symbolic processing cannot be addressed; figures of sensing cannot be configured with a Receiver.
Here there is a subtle difference between sensing and saying in their grammatical realisation by verbal and mental clauses that project reports. In the verbal case, the Receiver is a grammatical constituent of the verbal clause, as in she told him —> to leave. Being a constituent participant, it can serve as Subject in the verbal clause: he was told (by her) —> to leave. However, although the mental case, e.g. she wanted him to leave, looks similar, it is not, since mental clauses do not have a Receiver. The element him is a constituent of the reported clause: she wanted —> him to leave. Consequently, it cannot serve as Subject in the reporting mental clause; we cannot get He was wanted (by her) to leave. The difference also shows up clearly when the projected clause is in the passive voice; contrast the incongruence of she told the car to be washed with the acceptability of she wanted the car to be washed. The analysis shows the difference: she told the car —> to be washed (where the car gets interpreted as Receiver) : she wanted —> the car to be washed

∞ 

¹ As always, there are departures from the general principle. For instance, while 
she asked/told/persuaded/implored/encouraged/promised/threatened the car to be washed 
is incongruent, 
she ordered the car to be washed 
is perfectly fine, which suggests the mental model of the constituency boundary between the projecting and the projected clause — he ordered —> the car to be washed. This gets support from the fact that the car was ordered to be washed is odd.