A SYNOPSIS OF ROLE AND REFERENCE GRAMMAR 127
In terms of the RRG linking theory, the theory of control determines which
core argument must be shared between the two cores.
The linking from (114a) to (113a) may be represented as in Figure 41;
the numbers correspond to the steps in (115). The linking of each of the LSs
proceeds as in simple sentences until step 4. At this point, a potential prob
lem arises, namely, there is no argument slot in the linked core for the actor
of the embedded LS to link to, and if it is not realized in the syntax, then
the CC will be violated. However, because it and the actor of the matrix LS
are identical, they can both be mapped into the same syntactic function.
This actor-actor identity follows the theory of control, since try is neither a
causative nor a directive speech act verb. The "same-subject condition" on
the complement of try follows from the CC. If the LS being linked were try'
(John, [[do' (Mary)] CAUSE [BECOME clean' (clock)]]), there would be
no way to realize Mary, the actor of clean, and therefore the CC would be
violated. Hence there is no need for any kind of "same-subject" stipulation.
The theory of control plays a vital role in linking from syntax to seman
tics when the verb in the matrix clause is transitive. The procedure from
(70) is repeated below, with the addition of a step involving control.
