With this thesis, we have offered a physical theory of language that, we think, can withstand the scrutiny of scientific inquiry. We have also set out to establish the usefulness of a physics model with which to talk about the changes that occur in language evolution resulting from linguistic transactions.
We have presented the need for such an approach by establishing why a semantic theory is inadequate in describing most of language. We have also established that there are precedents for a physical theory of language in the research of several linguists and cognitive scientists such as Derek Bickerton and William H. Calvin.
In the effort of capturing salient features in language evolution it became necessary to introduce new vocabulary, such as attenuation and borrowed vocabulary such as phase transition. We have used these notions to describe the erosion of perceptual cues that correlates the use of vocables, erosion that, pushed to a critical point, will transform some lexical vocabulary into functional vocabulary.
In conducting this research, it became apparent that the small-scale dynamics of change that can be observed in the history of vocables could be mapped onto much larger-scale dynamics of change such as the changes that could bring about the phenomenon of grammaticalization.
Jennings' theory of connective vocabulary was our starting point and the inspiration for this research. Jennings has described how connective vocabulary is vulnerable to misunderstandings in its scope arrangements. These misconstruals may lead to mutations that give rise to new uses in functional vocabulary. He also establishes the ancestry of functional vocabulary in lexical roots and explains how these findings highlights the inadequacy of a semantic theory to describe this vocabulary. We then supposed that the dynamics - attenuation - that brings about functional vocabulary is also responsible for many changes in the use of all vocables. Moreover, that attenuation is exploited toward the propagation of linguistic systems. This assumption has led to the introduction of some physics concepts, the first one being efficiency.
The concept of efficiency was introduced as a driving force in the dynamics of language evolution. Efficiency is a phenomenon that many living systems are governed by. We have described how generally we use what we already have for novel tasks rather than generate new tools for novel situations. We have described how people, in their use of language, use a similar strategy. For reasons of efficiency, language becomes transformed with its use in a population. The dynamics of shareability extend vocables beyond original perceptual cues while still generating specific effects.
The transformation of vocables involves several dynamics, many of which can be described using other physical constructs. We have borrowed terms from classical and statistical mechanics and demonstrated a direct correlation with the dynamics of language evolution. We have used the concepts of propagation, dispersion, elasticity, degeneracy and phase transition to illustrate how vocables become attenuated. We suggested that attenuation in vocables will occur as a consequence of a negotiation of context between the members of a group, hence generating coherence for the use of vocables. This negotiating process involves dynamics that can be described using the physical concepts that we have mentioned previously. In order to formalize our approach we have used modeling, particularly the modeling of phase transitions.
In an effort to support our claims we have referred to the work of several researchers to show how models of phase transition can describe successfully certain dynamics characteristic of many phenomena that can be found in biology, economics and physics.
We then have focused on the description of a standard Ising model. We also introduce our modified version. We have demonstrated that changes in language can be illustrated by a simple, mathematically tractable model such as the Ising model. Though somewhat limited, it is relevant to use such a model to explore further aspects of language, such as first-order phase transitions, that are not so easily historically tractable. We have suggested that language can be thought of as a system in quasi- equilibrium, a system that changes, however, slowly enough to allow a population to share it. Finally, we have shown that our heuristic approach to the Ising model was viable and that our results prompted us to explore more sophisticated models, compatible with our thesis.
Formal scientific inquiry requires of a theory that it be compatible with a larger body of knowledge. We have shown that our thesis is compatible with a microscopic account of language given at the neural level.
We called upon the work of William H. Calvin to help us in this task. We have found that our account of linguistic transaction is compatible with his microscopic representation account based on neural firing pattern replication. We have suggested that functional vocabulary can result from certain kinds of neural pattern replication or partial replication. Calvin also uses the concept of phase transition to describe the resolution of competition between synaptic structures. Additionally, we suggested that Calvin's model could benefit from our approach by including additional concepts such as percolation. This could explain the emergence of stable patterns in low connectivity, patterns that could account for some kinds of generalizations and perhaps certain types of vocabulary such as functional vocabulary.
The requirements of a successful theory requires that our approach can be applied to a macroscopic view of language, as well, so we have tested our theory for the context of the emergence of language. Most of our theory has based its premises on a diachronic account of vocabulary. But our theory has also accounted for a time before vocables, when the dynamics we have describes may have been present in the soundscape of hominids and how they might have led to proto-language. We have pondered on how naturally occurring sounds and vocalizations may have been slowly exploited toward prosodic features of language. We also have elaborated on the role of prosody and how functionalization of some vocabulary may play a role in preserving prosodic flow.
From a generalized concept of functionalization, we have supposed a transference in priority, from visual cues, towards auditory cues. This possible ancestral dynamics may have led to shared communication systems. We suppose that vowels may be the descendent of natural vocalization while consonance may be the result of vocally reproducing rhythmic punctations that occurs with say, hitting a surface.
In closing, we have claimed that our approach is universal by suggesting that the concept of attenuation can correlate a macroscopic view of the emergence of proto-language in a population and microscopic view of this phenomenon, in neural terms.