Tone of Voice and Mind : The Connections between Intonation ...

Chapter 8
A bilateral neural network simulation
Synopsis
A technique for the bilateral activation of neural nets that leads to a
functional asymmetry of two simulated “cerebral hemispheres” is described.
The simulation is designed to perform object recognition, while exhibiting
characteristics typical of human cognition – specifically, a dual focus of
attention, corresponding to the “nucleus” and “fringe.” Sensory neural nets
self-organize and the system is then taught arbitrary symbolic labels for a
small number of similar sensory stimuli. Mutual homotopic inhibition across
the “corpus callosum” produces functional cerebral asymmetries, i.e.,
complementary activation of homologous maps within a common focus of
attention – a “nucleus” in the left hemisphere and a “fringe” in the right
hemisphere. An object is recognized as corresponding to a known label when
the total activation of both hemispheres (nucleus plus fringe) is strongest for
that label. The functional dualities of the cerebral hemispheres are discussed
in light of the nucleus/fringe asymmetry.
A distinction was drawn in Chapters 6 and 7 between consciousness and cognition
in terms of basic properties of the neuron. In line with that view, it is unrealistic
to believe that a silicon-based computer could become conscious in the
normal biological sense of that word, because it is specificallytheexchangeof
materials across the cell membrane that gives the neuron and neuron-systems
their property of sensitivity to an external world. For the neuron, “receptivity to
the surrounding world” is not a metaphor, but rather a factual statement about
how it learns of external changes. For existing silicon systems, “receptivity” is
metaphorical and all knowledge of the external world is hard-won through data
processing. This fundamental difference between the living and the non-living
worlds implies deep and probably lasting differences in the quality of “minds”
of men and machines.
Nevertheless, the information-processing that neurons are capable of can
certainly be simulated (but not the feeling of the information-processing!), so
that, in principle, any form of cognition that can be explicitly defined can be
simulated in a computer. Thus far, many sophisticated simulations have been
developed world-wide (e.g., Edelman 1992), but the issue of the bilaterality of
the nervous system, particularly with regard to language, has been thoroughly