Perceptual shift as paradigm shift

Perceptual shift as paradigm shift

In order to think about active inference at the level of the mortal substrate or good biomimesis, it helps if we can talk plainly. We can think of active inference as a cognitive function as something that perhaps Sherlock Holmes might be good at, but insight rather than deduction. Let’s reserve some open space about what that really means but we do understand from the free energy principle that the function of active inference, predictive coding, belief updating, is basically homeostatic.

This accords well with how we understand both interoception and proprioception, or physiological and physical homeostasis respectively. Physical homeostasis occupies a special position in the model, where it is located at the core. This spatial metaphor affords easy anatomical identification and the system, being the physical axial line, is clearly a functional unity. Moreover the axial physical system is much simpler and more straightforward than the physiological system. It both affords simpler and deeper insights.

Active inference as a sensory motor perceptual system is experienced as proprioception, or physical homeostasis. This system possesses anatomical biomechanics that are clearly evident from its structure. The expression of active inference in this system is very easy to describe mechanistically. It is also very easy to identify kinesthetically. We have rich sensory awareness in the form of the muscle sense. It is also clear that there is a very meaningful cognitive metastability that is developed in the neuraxis.

The postural system is mechanistic but assuredly is not merely mechanical. The postural function of the core system is supported by the cerebellum, the massive human hindbrain, a pattern recognition engine. The cerebellum is a powerful computational system containing more neurons then the rest of the brain, including those parts that are responsible for our conscious awareness, physical prowess, and rational intelligence.

That the cerebellum is a very distinct part of the brain, a second or “mini” brain, is evident by its location. The cerebellum is separated from the rest of the brain by leathery folds of “dura mater” called the cerebellar tentorum. We can think of this membrane as the outer boundary of the inner model, a roof over the animal intelligence, shielding it from the rest of our activity. In this way it has a very special relationship with the rest of the brain, funnelled through two pairs of basal nuclei. In this way it is able to maintain a very distinct self identity. It is very much like a second brain. It is the location of what we can refer to as our animal intelligence.

Active Inference as a sensory motor system, or as physical homeostasis, is a plain and simple (self evident) anatomical system in the body because it is the deep axial core. It is one of those things that becomes obvious once you know to look for. Conscious recognition requires us to be able to name it and we now have the scientific and conceptual capacity to do that. The active inference framework fills in a lot of the blanks.

What is important to reinforce however is the idea that active inference as physical homeostasis is not the same kind of willful, volitional adaptive behaviors and associated habits of mind that we are familiar with. The adaptations we cause with the intellect are very different from those that are the result of cerebellar computations, with its muscle sense and self-referentiality. The wiring is entirely different and results in long continuous loops rather than the linear narrative semantics of exteroception. We are aware of this as flow, whose the teleological causal dynamics are self-evidencing rather than other evidencing.

This form of cognition is valuable because we do not merely predict the outcomes of our actions but we foretell them. Active inference as a sensory motor system (in our body) is something that emerges from the core of the generative model as it is experienced in our guts. We now possess the language necessary to explain this part of our mental experience in a way that is not only physically accurate but meaningfully self evident. This deep and self-directed, self-aware system is qualifiable but not quantifiable, while in contrast the surface system is quantifiable but not qualifiable. That we come to possess both as complementary tools is the paradigm shift.

You won’t really understand what active inference means if you are not able to perform this kind of perceptual code shifting, performed physically, haptically, mechanotransductively, in the muscle sense.

This is our animal intelligence and we need it now more than ever.