#3 Neocybernetics = feedback control theory + feedforward command theory

The term 'cybernetics' usually refers to computers and computation in the most generic way. This is unfortunate because that is not its correct meaning. Originally, it referred exclusively to the science of 'systems control'. Dedicated devices which embodied these principles were called  'control systems'. There are issues of ambiguity around the term 'control' which are largely avoided by using the alternative terms 'governance', 'guidance' or 'direction'. In this discussion, feedforward governance (adjective 'gubernatorial') is called 'command' while feedback governance is called 'control'. 

Powers' identifies six levels of PCT in animals including humans. Because each level employs perceptual common coding (PCC), force computations are always kept local to each level, and hence remain unconscious. Reflexes, ie sets of stimuli followed by related responses, typify this kind of process. PCT therefore explains the multi-level nature of consciousness in a very neat, non-ad-hoc manner. Each level generates a conscious signal whose magnitude is roughly proportional to the deviation of that levels sensory inputs from each of its respective matching setpoints. Thus, consciousness changes its focus according to which level/s exhibit/s greatest cybernetic deviation/s.

There is a problem with PCT and it concerns its reliance on feedback as the primary system dynamic. Because this assumption is actually one of PCT's original axioms, it is not easily modified. Unfortunately, feedback alone fails to adequately describe voluntary behaviour, which is by definition a feedforward (command, leading) signal type.

Animate governance systems have another class of setpoints which are not bounded in the same way as those that anchor conventional structure-preserving, static feedback mechanisms. This second group of setpoints function not to preserve static (ie structural and metastructural) values, but to disturb them according to deliberate (ie voluntary) criteria. 

We have seen how the first, multi-level static setpoint system naturally generates multi-dimensional conscious states (percept fields). We must augment this model with the second, multi-stage heterostatic setpoint system, which generates multi-scaled voluntary states (emotional gradients). 

We saw in the previous section how Libet's failure to consider volition as well as perception as parts of consciousness led to epistemological chaos. This epistemological problem is finally remedied by expanding the abstract ontology of internal states of complex systems in a substrate-independent way. 

This second system can be termed 'pseudodynamic' because it uses slow transitions between sets of static states to generate a computable dynamic behavioural framework. This second system was codiscovered by Anatol Feldman and this author (M.C.Dyer). Feldman's research is focussed on the detailed implications of the discovery to the PNS and related musculoskeletal performance. Dyer uses his term 'neocybernetics' to describe the discovery which Feldman calls 'Equilibrium Point Theory' (EPT). Neocybernetics is a more generic, abstract systems approach to the topic than EPT. See Figure 3 below. 

Neocybernetics represents an innovation in science because it allows finite automata theory (ie state-transition mechanisms and logic) to be used to realistically model biological computations.

Figure 3a. The diagram above demonstrates how the neocybernetic paradigm may be mapped onto the GOLEM two-channel model brain architecture. Static setpoints are termed 'biases' that are by nature 'tonic', ie possess values that are slowly varying. Heterostatic/pseudodynamic setpoints are termed 'offsets' that are by nature 'phasic', ie possess values that are quickly varying.