Tool making can provide a metaphor of an evolutionary account needed in the description of the dynamics of language evolution. Without adopting the language of intention that is usually involved in the description of tool making, we describe evolutionary dynamics that are mirrored in language evolution. Tool development is an evolutionary process that complies with architectural constraint inherent to any evolutionary process. These evolutionary dynamics involve concepts of energy conservation and propagation. The way to understand these concepts is in considering the physical effects that they yield. The story emphasizes the subtle balance between using what is already there and generating novelty in the context of increasingly various tasks.
The invention of the Swiss Army knife might have come about in a situation in which several different tasks require an array of dissimilar tools. However, it may be that only bladed tools are available. One-tool-for-one-task is a high energy state because it requires a redesign and remanufacture as a new task emerges. However, the energy put into developing the tool pays off in the coherence - the focus the new tool brings - of the work it was designed for. Some of the energy invested in the development stage is recouped in the ease to perform the task since the tool is perfectly designed for that task. This is a stable state, as a minimum is reached between the input and output of energy.
On the other hand, there is an increasing variety in the tasks to be performed. New tools are constantly required. There is a critical point at which the energy recouped in the ease of performing the task is not worth the energy investment required at the new tool development stage. What to do then?
So what if tool availability is restricted to a bladed tool - a knife - with a handle, an edge, a tip, a width and a length? We can use these features in developing a new tool. The width of the knife can be increased and a depression can be added to perform a scooping - edge forward - action. Hence is born the spoon. As cutting and scooping tasks are encountered it may be useful to have both edged and scooping features on hand as to save time and energy. Most of us are familiar with sporks, which are both a spoon and a fork in one, with which one can both spear and scoop. However, the ease with which it can do both is questionable. It may be more useful to add an additional tool to one handle than to modify one tool to perform several tasks, but we will come back to this dilemma later.
A strategy to add a spoon to a knife handle can be adopted. The spoon features can then be refined while retaining the use of the knife. The work that the new tools - or collection of tool - can do compared to the work of the knife alone has been increased without investing too much effort into it. However the handle is now a little larger, the knife and spoon may interfere with each other slightly and the complete tool is less reliable. The consequence is an increase in work for a minimal energy investment but the effects are not as specific as those originally occasioned. Originally the effort in performing a cutting task is less because the knife is designed for that purpose. The balance of the handle and the weight of the blade is in perfect harmony, so as to aid the cutting strokes. The work done is highly coherent. Now that the knife is not quite as steady because the handle is slightly bigger and does not quite fit the hand, more effort is needed to perform cutting. The coherence of work done is decreased. As more extensions, such as screw drivers and bottle openers are added to the handle, and, suppose, the size of the collection is reduced to increase portability, less effort is required in manufacturing and carrying the collection of tools, compared to creating and carrying new individual tools. However, there is an increase in effort in using this collection of tools. Eventually, each new extension becomes a problem for two major reasons:
In the next stage of evolution individual tasks are evaluated to see which ones are repeated more often. For these tasks it may be more economical to invest the effort in making a specific tool. However there is no need to discard the all purpose tool, it can simply be put in a support role for the specialized tools. With a combination of specialized tools and an all purpose tool most situations can be tackled effectively. As such, a fairly high level of coherence can be achieved.
The dynamics involved in language evolution follows a similar pattern. Language rich in perceptual associations is a specialized tool. As these vocables are generalized to do more work, the specificity of effects that are generated by them is less coherent. Once vocabulary has lost some of its perceptual cues and starts generating incoherent effects, it can be further exploited by gaining a support role for linguistic instruments that do generate specific effects.
We think that functional vocabulary has evolved to support vocabulary on its way to become increasingly less specific but lexical nonetheless. Functional vocabulary has lost most of its perceptual associations and seems to be restricted to a syntactic role. The one purpose vocabulary - lexical even indexical - is a higher energy state and highly coherent because it generates specific effects. It is easily accessible for it is entrenched in perceptual cues, but as such, requires much neural space. For example: If we refer to THAT table in THIS room the effects will be very specific and will be entrenched in very specific perceptual cues. The all-purpose vocabulary - functional - is a low energy state. We know how to use it despite not having a semantic theory for it. It is fairly coherent in the effects that it generates within its syntactic role. However, it does not generate very specific effects in relationship to perceptual cues. Jennings has supplied many examples to that affect, but, for example is very elusive and it is very difficult to associate it to any experiential cues.