Re-Thinking the Place of Semiotics in Psychology... 125that the speech warnings are the most easily learned and recalled (as in the case of the signal
―unknown‖ to indicate the target referent "unknown threat").
Clearly, this second (triadically conceived) definition of signal-referent relation strength
entails a different, but much less fixed classification of sounds, for it all depends on how well
learned are the signal-referent connections, and on the context of their occurrence. The price
of completing the triadic signifying relation by including the cognitive contribution of the
person is inevitable variation and uncertainty in classification. Ironically, however, this does
not matter. For it is clear that testable predictions regarding ease of learning are vacuous,
because they are already contained in the classification, such that the dependent and
independent variables cannot be separated.
In summary, the standard method in this research area, that of constructing and testing
classification systems, conflates two different understandings of the notion of signal-referent
relation strength, one based on a dyadic relation between signal and referent, the other based
on the full triadic relation between signal, referent, and person. However, when these two are
disentangled, it becomes clear that the conflation disguises serious problems which mean that
neither is a useful way for this field to advance. Predictions derived from taxonomies based
on the dyadic relation will produce results which are inevitably confounded by learning.
Predictions derived from taxonomies based on the triadic relation are vacuous, for they are
already answered in the classification itself, rendering pointless their testing. The implications
for auditory warning design are that the prior knowledge (the "cognitive contribution") of the
user must be taken into account earlier in the design process.
Now, when the third term of the signifying relation, the person or cognising organism, is
properly taken into account, it becomes clear that cognition will not be the only relevant
aspect. For example, in the auditory warning literature, neglect of the prior learning of the
person has prevented critical scrutiny of the very notion of the "abstractness" of a sound (such
as a siren to indicate an approaching ambulance or applause to indicate approval). Yet there is
a substantial body of literature relevant to this issue. It deals with the way our reactions to
characteristics of sounds and symbols have been shaped by the combination of evolution,
aspects of our physiology, our bodily experiences, and the natural environment. This takes us
to the next contribution of semiotics-psychology integration.
Using Iconicity as the Bridge between Conceptual Metaphor and
Nonconventional Symbolic Phenomena
I have referred to the role of the cognitive contribution (viz. prior learning) of the user.
However, since the cognising term of the cognitive (and, thereby, signifying) relation is an
embodied, motivated subject, it would be surprising if bodily, motivational and emotional
experiences were not also important in shaping our ways of thinking. In psychology this is
gradually being acknowledged in the new approaches to "embodied" and "hot" cognition,
which draw upon earlier theories of conceptual metaphor, although this work remains
relatively isolated from experimental research areas in which it could usefully be applied. By
contrast, nonconventional symbolic phenomena, which occupy a shaky position even in
semiotic classification, have been relegated by mainstream psychology to the "too hard"
basket, and receive barely any attention at all. Yet, as I shall argue, the Peircean concept of
iconicity can be used as a bridge between the two. In this way, semiotics-psychology