Information Technology Reference
In-Depth Information
neural network with such characteristics as, new neuron is born, neuron grows by
extending its dendrite, a neurons makes links to the others by touching its den-
drite to the axon of the others resulting in a neural-network, every neuron learns
from stimulus from external world and the success or the failure of its response.
If further more it is shown that a neuron could behave like a logical inference by
learning, then these biological characteristics can be a latent Universal Grammar
and the created neural networks represent a language expression. The difference
between ordinary operation of neural network like recognition and this logical
operation is made clear.
By showing that these concepts can be connected to the vocal organs, and
also to sensor organs by means also of purely biological ways, a possibility of
originating language could be explained.
This was the simplest holistic language. It was shown that it could develop
to social language and to compositional language in the same framework. After
some time this language was symbolized. Symbolization was a next big evolu-
tional stage of language. It depends deeply on biological capability of memoriz-
ing. But this function is not made clear yet. In order to analyze the process of
symbolization we must wait the researches from neuroscience and brain science
on this matter.
