Introduction
We are trying to investigate the role of language in evolution. We are starting with the assumption that much of evolution revolves about physical action, whether for feeding, reproduction, or attack and defense. Another related thread is based on the assumption that learning is crucial for action related to evolution. We therefore want to look at the role of language in action and in learning, for many species, not just humans.
Our investigation is based primarily on human actions on a stage. Action in the world is represented by the movement of one or more actors. Primarily the actor moves in accordance with a script either from detailed choreography or from very detailed and explicit stage directions. We assume that he has memorized and practiced this script, which becomes his plan for action. The actor will also be using information from his eyes and his ears to provide positioning and timing information. Our investigation focuses not on the diversity of action across species and habitats but rather speculates about inner mechanisms that are not open to inspection or to empirical measurements. Instead we shall a model with some minimal competence to generate the actions and construct arguments that the actual mechanisms in the brain must function with partial equivalence to the mechanisms in our model. In other words, a better theoretical model, to compete with and replace our model, must be more competent to generate the actions � it must be better at simulating the actions, the timing, and the conditions modifying the action. Rather than depending on the language of mathematics to specify our predictive model, we shall use the language of programming and computer simulation. To separate components of our model from implementation details, we shall use systems analysis a la von Bertalanffy.
Action is described in terms of the movement of the actor�s bones � representing overall body movements. Body movements (and by inference bone movements) are public, i.e. observable and measurable. They are visible by the actor himself, by other actors, and by outside observers. They are recordable on video. The neural activity that leads to the action is not specifically and directly measurable, not even by a functional MRI. In our investigation we therefore need to differentiate between public and private information.
Visual perception is described in terms of video, with its 2 dimensional rasterized representation with colour depth or gray tones. The flow of time is represented by successive video frames. To simplify and synchronize components of the model, the same frame by frame sequence concept is used to represent time for the action.
For this model, again for the sake of simplification, we only look at overall body movement. Hands, speech production, chewing, and many other actions are either treated separately or ignored. Only visual perception is used. Other senses, as well as proprioception, is either treated separately or ignored. Many other major and minor simplifications are made to make the programming easier. Some of them will be discussed as they arise.
What drives the action?
We envisage 3 distinct types of information leading to action. The first, most obvious is spatial information to give direction to the action and avoid obstacles. The second is a plan that sets out the steps that need to be followed to make the action successful and accomplish the goals. It is through the plan that we hypothesize that the rudiments of language have contributed to evolution, but more about plans and language later. In the initial chapters we will explore the relationship between perception and action which will constitute the setting for our subsequent investigation. Finally, we should not forget about other senses and proprioception which could also play a part in action, but will not be part of our investigation.
Visual perception brings up the issue of information exchange with others. Actors can perceive public information but do not have access to other�s private information. Even access to one�s own private information is severely limited, with concepts such as consciousness. For example, for access to our own internal information flows, we just do not have access to the nerve impulses being sent to particular muscles around a specific joint. However, we can visually perceive when the other actor has reached the chair on the stage and sat down � which might be our clue to perform another action � such as walking toward him.
Public information exchange
Generalizing to other species
The action component generally deals with moving the bones of the skeleton to accomplish the action sequence. Most species have similar components of neural connections, muscles, bones, and joint rotations to achieve the desired action sequence.
The spatial perception component generally deals with recognizing objects in the spatial environment and with locating the actor relative to these objects. Usually this perception is based on vision, but not always. Most species have similar needs to recognize objects, as potential food or as danger.
The goals and plans component deals with what the individual (the actor) needs to do, and with what the appropriate conditions might be to accomplish the action plan. Some action plans might be genetic, such as imprinting for chicks, and some may be learned (e.g. through play and an apprenticeship-like process).
What about language and evolution
We are starting with the assumption that animals somewhat higher in the evolutionary scale have the same requirements for managing their actions. We hypothesize that they share the basic requirements for information processing components we have called plans, spatial perception, and action through bone movements. The initial part of our investigation will stay focused on these components. At the end of this initial part of the investigation we shall focus on the role basic language mechanisms might play in the information flow, the information encoding, and the information transformations required for systematic action. It is only after this part is completed that we will look at the role language might play in allowing us to edit action plans, perceptions, and predictions. Language for signaling and communication does not come till much later.
In all of this we would hypothesize that the basic information processing mechanisms are shared with a lot of animals, but that editing requires somewhat higher skills that appear somewhat later in the evolutionary sequence. Communication, signaling, and �shared reality� come much later. We anticipate, and shall look for, many more stages between the basic mechanisms and full communication.