Alternative paradigms for investigating the evolution of skills

We start with paradigms for investigating skills, and then discuss paradigms for investigating the evolution of skills.

Four paradigms for investigating skills and the evolution of skills:

The first paradigm relies on observations to collect data. The paradigm is to collect thousands of skills through observations of consistent behaviour patters in a variety of species, including closely related ones. We could isolate evolving patterns in closely related species, and try to pinpoint emergent patterns. Lab experimentation is difficult because of the time spacing between successive generations of the type of advanced vertebrate species that are of interest. There may be opportunities to use space instead of time to track the evolution of skills. (Ring species are an example.)

The second paradigm is to build neuronal models of selected skills, and to try to make them functionally complete, i.e. working prototypes.

The third paradigm is to detect DNA sequences of selected innate skills and to look for the correspondence between the variations in the sequence and variations in the behaviour patterns. We would use comparisons of DNA sequences of closely related species with known skill differences.

The fourth paradigm, the one utilized for this research, is to treat a skill as a set of stored instructions, analogous to a computer program. When activated, these instructions should produce the behaviour patterns associated with the skill. Following the analogy in the examples, we are constructing an instruction manual for the skill.

• This approach overlaps with the neuronal modelling approach, but does not commit to any specific hardware - software architecture, since any abstract set of stored instructions can be implemented as dedicated hardware, as software running on a general processor, or as hybrid.

The fifth paradigm, also utilized for this research, is to treat evolution as an information-processing mechanism (as commonly done in bio-informatics). Rather than focusing on attributes with relatively few dimensions, we treat the set of stored instructions representing a single skill as a complex, multi-dimensional attribute that evolves over generations.

• meiosis, i.e. combining the father's and mother's skills is an interesting challenge
• mutation of innate skills is another interesting challenge

Combining models of skills with models of the evolution of skills:

Using the above paradigms, an information processing model of a skill is an abstract set of instructions that generate actions and that collect information from perception to modify the action. An information processing model of the evolution of the skill is a program that works on the abstract instruction-set representing the skill.