- The current research investigates the evolution of skills (in vertebrates). We suggest looking at some examples, starting with an innate skill, a learned skill, and a group-level set of complementary & interactive skills. We also suggest looking at the potential timeline of emerging skills, as well as considering some hypotheses, and some alternative methodologies and paradigms.
        - What does it take to walk? We start with the action, then look at the skill, and then at the movement. We propose an innate skill-automaton.
        - How do we stop from falling over while walking up a slope? The skill integrate information from the perception of gravity, and adjusts joint angles so that the plumb line from the centre of gravity to the floor does not go too far from the supporting limbs.
        - How do we intercept someone or chase prey? The skill needs to utilize perception actively to control the direction of the chase. The skill needs to twist the hip-joint to face toward and run toward the prey. The skill needs to integrate conditions of the form "if ... then ...", or "unless ... do ...".
        - How do we mimick someone, such as for imprinting or following a fitness leader? Information from visual observation is converted into joint-angle instructions for action.
        - How do we learn from mimicry? Some skill needs to abstract and store the output information from an innate skill and integrate it into an internal representation of a learned skill in order to generate appropriate future action. In language terms, the description of the leader's action has to be stored while preserving the time sequence: "first do ... then do ... then do ...".
        - How do we optimize the chase by running in a straight line to intercept the prey? The skill needs to predict where the prey will run and where the predator might intercept. The direction for the chase is chosen from the predicted point of interception. In language terms, since the prediction has to precede the action, we need hypothetical constructions such as: "if the prey were to run in this direction and if I were to run in that direction then I could catch the prey at ..."
        - How do we coordinate action with group-level complementary skill sets. From simple information exchanges to communication - this may be the root of spoken and written language for humans.

The research is very much a work in progress, and is documented (at least in part) on this Web site using the organization below:

• An abstract and general discussion of the research paradigm:
• single skills:
• Action components: (output)
• Atomic action components of a skill
• Molecular components: a frame, minimal action separate from consequences
• Assembling frames from atomic & split-frame components (e.g. for meiosis or new actions)
• the relative geometries of joint angles
• Experiments with split-frame components and timing (left-right, top-bottom)
• Assembling more complex actions by glueing molecular components together
• Inner information processing
• integrating perceptual information to modify (select or edit) action components
• Adapting the action rather than separate and almost duplicate skills
• requirements: catching prey, mimicry, and in general utilizing perception
• the internal logic of skills: conditions, invocation, prediction, planning
• The perception (input) component of a single skill
• gravity
• visual direction for chasing etc.
• perceptual discrimination and minimal information content (in bits)
• visual features and mimicry
• TOC including an abstract discussion of skills:
• Group-level skill-sets for collective behaviour
• Evolution:
  • TOC for attribute-based evolution (pre-skill)
  • TOC
  • combining skills from both parents, evolving skills and skill-sets
  • History of the research: A top-level TOC provides more of an overview to prior versions, set in a framework. It is embedded in a wider story with a history, context, and speculations. One such speculation is that an inner language (proposed for representing & compressing skills, and for supporting prediction and planning) might, in future, help to explain human consciousness.
  • (list of Web pages)