[Search Blog Here. Index-tags are found on the bottom of the left column.]
[Central Entry Directory]
[Computation Entry Directory]
Neuronal Assemblages and Reassemblages
in
Hans Flohr
"Qualia and Brain Process"
Emergence or Reduction?
Essays on the Prospects of Nonreductive Physicalism.
Activity in our brains neural networks "is coupled with reorganization of these nets." *(225c)
Flohr's hypothesis is that "the occurrence of phenomenal states depends on the rate at which activity-dependent synaptic changes occur and neural assemblies are formed." (22cd)
Assemblies
In The Organization of Behavior, D.O. Hebb proposes that
repeated stimulation of specific receptors will lead slowly to the formation of an 'assembly' of association area cells which can act briefly as a closed system after stimulation has ceased; this prolongs the time during which structural changes of learning occur. (Hebb, qtd in Flohr 225d)
Later in 1959 he adds:
The key conception is that of the cell assembly, a brain process which corresponds to a particular sensory event, or a common aspect of a number of sensory events. This assembly is a closed system in which activity can 'reverberate' and thus continue after the sensory event which has started it has ceased. Also, one assembly will form connections with others, and it may therefore be made active by one of them in the total absence of the adequate stimulus. In short, the assembly activity is the simplest case of an image or an idea: a representative process." (Hebb, qtd in Flohr 226a)
Neural nets can self-organize on account of plastic synapses, called Hebb synapses. According to Hebb,
synapses on a neuron that are active while the neuron discharges will be strengthened, whereas inactive synapses will be weakened. Synapses from different inputs that are active at the same time on the same neuron will be reinforced and selected over others. (226)
Self-organized assemblies emerge from random beginnings.
When applied to nets of spatially distributed groups of neurons with non-specific, random interconnections, these so-called Hebb rules lead to a relative stabilization and association of neurons firing in a correlated fashion. An assembly of preferentially connected, coherently active cells is formed. If coincident activity is induced in some neurons of such a net by a patterned input, an assembly will be formed because the synchronous activity selectively modulates the pathways connecting these neurons.
The assembly detects and encodes the coherent properties of the stimulus pattern so that a representation of that pattern is generated. Once the assemblies have been formed, they would function as detectors of the same or similar input patterns expressing the detections of coherent features by coordinating their activities. It is easy to envisage that the output of such assemblies could in turn be used as input to other modifiable nets which then would reorganize their structure as a function of this input. Iteration of such processes would generate more and more abstract metarepresentations. (226c)
No comments:
Post a Comment