Participant: Terrence W. Deacon
Affiliation: Department of Anthropology, U.C. Berkeley
Format: Plenary Speech

Complex (adjective) — etymology: from com- “with” + plex from plectere orplicare “to weave, braid, twine, fold.”

Why are computers so radically different than brains in terms of the presence or absence of intrinsic phenomenology? The difference is one of complexity, but not complexity in mere numbers of elements, interactions, operations per time and space, or even logical depth. The difference is dynamical.

In better agreement with the etymological derivation of ‘complexity’ than the contemporary colloquial definition, I propose a measure of the complexity of a system that is largely orthogonal to what can be collectively described as mereological, compositional, or information theoretic conceptions of complexity. These latter measures are based on some means of quantifying numbers of distinguishable and analytically irreducible components and their interrelationships, or some quantification of the analytic difficulty or computational size of a system’s description. In contrast, I propose a measure that captures the degree of convolutedness or recursive infolding of dynamical relationships upon themselves that results in distinguishable emergent levels of temporally asymmetric system attractor dynamics; i.e. discrete nested inversions of orthograde dynamical tendencies.

A system with greater dynamical depth than another consists in a greater number of nested emergent dynamical levels. Thus a thermodynamic system has less dynamical depth than a morphodynamic (e.g. self-organizing)system has less dynamical depth than a teleodynamic (e.g. living or mental) system, and so forth. Dynamical depth can provide a precise and systematic account of the fundamental difference between computation (low dynamical depth) and cognition (high dynamical depth), or inorganic chemistry (low dynamical depth) and living chemistry (high dynamical depth). Systems with low dynamical depth may consist of many more components and interrelationships between these components than systems with high dynamical depth. Dynamical depth is essential to explain the degree of agency, behavioral autonomy, internal coherence, semiotic capacity, and sentience of a system whereas mereological, componential, or information theoretic measures of complexity provide no insight into these phenomena.