Open-Ended Elaborations in Creative Metaphor - Computational Creativity Research: Towards Creative Machines - page 223

Note also that in all this, the only view- specific correspondences used are the
very basic one of relative vertical position relationships in pretence corresponding
to control relationships in reality, and the identity-like correspondences for the con-
versations and managers. Thus, the meaning arising from (3) is generated mainly
by VNMAs, not by view-specific correspondences. Work on the ATT-Meta approach
indicates that metaphorical utterances often get much, and in some cases all or almost
all, of their effect via VNMAs rather than directly from view-specific mappings. The
latter often merely provide a scaffold to allow VNMAs to handle the most important
information. The treatment of a wide variety of examples in [ 2 , 4 , 6 , 8 ] provides
evidence for this claim.
However, VNMAs are merely default rules, so their implications can be defeated
in specific circumstances by other evidence. This defeasibility applies also to the
effect of view-specific rules.
As another example of the use of VNMAs, there was no treatment above of the
effect of the “small” qualifier of the voice in example (7). Suppose this qualification
is added to (6), to get
(12) One small voice inside Mary was insisting that Mick was adorable.
The effect of the smallness is quite subtle, but at least some of it can be captured
by the notion that a small voice in a conversation is a relatively ineffective one (by
default). So, the voice, and therefore also the mental states of the subperson whose
voice it is, are relatively ineffective in the overall pretended scenario. Now, the sub-
person's motivational state corresponds under the Mind as Having Parts that are
Persons view to Mary's having a motive to believe that Mick is adorable. So, the
effectiveness of the subperson's motivational state in the pretended scenario corre-
sponds by the Causation-etc. VNMA to the effectiveness of that motivational state of
Mary's in the reality scenario. Moreover, as qualitative degrees to which situations
hold carry across by the Degree VNMA, a relatively small degree of effectiveness
in the pretended scenario maps to a relatively small degree of effectiveness in the
reality scenario.
Notice that the pretend subpersons themselves do not correspond to anything in
reality . It is merely that certain things about their mental states correspond to certain
things about the real person's mental states. This point leads into the next section.
11.3 The Anti-Analogy-Extension Thesis
In the ATT-Meta approach, as in conceptual metaphor theory, metaphor is based
on familiar analogies. An ATT-Meta metaphorical view is largely realized as a set
of entrenched correspondence rules, and VNMAs are additional correspondence
rules. 16 Nevertheless, a key point about the ATT-Meta approach can be called the
16 The realization of a viewcan also involve ancillary assumptions [ 8 ]. These are default assumptions
about the nature of the pretended scenarios involved in a metaphorical view. Figure 11.1 shows one