In order to understand the dynamics of language evolution we must inquire about the roles that linguistic interaction plays in human population.
Language is a binding process causing specific effects in a population. These effects must have a long history.
We share a linguistic system that binds us in many ways, one of which is a role of coordination in a population. Vocables are uttered and physical effects occur. A person can call up a friend to set up a meeting and gather in a particular place. With linguistic interactions, events may be set in motion that otherwise would be difficult. The coordination of activity occurs with some precision. But as explained in the Efficiency of Language chapter, the precision in linguistic interactions is only as refined as it needs to be in order to be shared in a population. However language does respond to the requirements of increased specificity by becoming more refined. If we assume that linguistic transactions play a significant part in coordinating human activity, which is not to say that language emerged out of a need to communicate, we may also assume that our non-linguistic ancestors may not have had the tools for the level of precision that humans experience with the help of linguistic transactions. The suggestion is that the level of precision, with which coordination occurs, will change as the communication tools in proto-human activity become more sophisticated. It may also be that the requirements, in activity, for coordination, applies adaptive pressure. The adaptation can occur because these transactions are coordinated as well. For example, in linguistic transactions, it is the specific permutations of words that make a sentence generate specific effects. Not just any combination of vocables will generate specific effects, however, vocables that belong to different categories seem to be required in order to form a sentence that is considered grammatical, that is, a sentence that can be understood by most users of the language.
There is a balance, it seems, between adaptability and effectiveness. Now, the role of functional vocabulary seems to have more to do with helping us in the construction of grammatical sentences than in generating very specific physical effects. Functionalized vocabulary is not the kind of vocabulary that we commonly use to make statements about the world; that is, as we will explain further on, its relationship to perceptual cues has been eroded beyond recognition. We mean this quite literally. Nonetheless, the stringing of vocables is done, we think, with some amount of precision. How much precision? As we have said, as much as is needed for the language to be shared in a population as the system propagates and as it is inherited by following generations. This means that it is not as precise as some may think. This lack of precision will allow mistakes in the production of language, mistakes that will be replicated and adopted in conversations. However, language is stable enough, in the short term, to be successful in its coordinative role and to be shared amongst a population.
At some point, our guess is that linguistic capacity was exploited toward the coordination of human activity, a role that might have overwhelmed any other use linguisticity may have had in its early times. The evolving capacity of human beings to coordinate their activities and the consequent changes in the depth and detail of actual coordination must itself have fostered profound, perhaps fundamental, changes. We know, for example, that we come from non-linguistic ancestors. Structurally, proto-humans were different from humans. The laryngeal morphology and vocal cords, the level of motor function involved with facial mobility and dexterity are some of the fundamental ways in which humans differ from proto-humans, yet, we have assumed that at some point in time there were human ancestors that did not have linguistic capabilities. We have no examples of the types of items that coordinated proto-human activity or the kinds of structural changes that promoted or sustained the slow emergence of language.
We must make assumptions about early language, based on recent language, and project back to earlier time. We assume that social organization involved types of communicational items, items that were the distant ancestors of vocables. We have examples of structural changes in the vocables of present languages, from which we can extrapolate, that is, we assume that the dynamics underlying contemporary linguistic changes are present throughout the history of structural changes in communicative behavior. We also have a theory as to how these changes promote a more detailed and versatile coordination of human activity. Though structural changes in linguistic types may occur at an accelerated rate, compared to the crawling rate of change in proto-human communication behavior, and may be less dramatic than the structural changes that promoted the earliest linguistic transactions in humans, they may nevertheless serve as approximate models of conjectured earlier developments in the evolution of language.