What Does ‘Information’ Mean in Integrated Information Theory?

Abstract

Integrated Information Theory (IIT) intends to provide a principled theoretical approach able to characterize consciousness both quantitatively and qualitatively. By starting off identifying the fundamental properties of experience itself, IIT develops a formal framework that relates those properties to the physical substratum of consciousness. One of the central features of ITT is the role that information plays in the theory. On the one hand, one of the self-evident truths about consciousness is that it is informative. On the other hand, mechanisms and systems of mechanics can contribute to consciousness only if they specify systems' intrinsic information. In this paper, we will conceptually analyze the notion of information underlying ITT. Following previous work on the matter, we will particularly argue that information within ITT should be understood in the light of a causal-manipulabilist view of information (López and Lombardi 2018), conforming to which information is an entity that must be involved in causal links in order to be precisely defined. Those causal links are brought to light by means of interventionist procedures following Woodward's and Pearl's version of the manipulability theories of causation.

New Classification of Collective Animal Behaviour as an Autonomous System

“…In summary, the combination of these three parameters models a situation where information transfer occurs only when the individual changes direction within their interaction radius, similar to existing flocking models [69,70]. Notably, using 'OR' to connect the parameters is illogical.…”

“…The manipulativist view [70] helps explain why IIT 3.0 is the best candidate for analysing those requirements. The causal-effect structure of an autonomous system represents 'what causes what', considering all counterfactual states [63].…”

“…The causal-effect structure of an autonomous system represents 'what causes what', considering all counterfactual states [63]. Lombardi et al correctly pointed out that the central idea of 'differences that make differences' represents intervention in the cause-effect structure of a system [70]. Note that this manipulation in the system does not need to be an actual one.…”

New Classification of Collective Animal Behaviour as an Autonomous System

“…As Tononi claims, "something can be said to exist only if it has cause-effect power" [61]; the causal-effect structure of an autonomous system represents "what causes what", considering all counterfactual states. Lombardi et al correctly pointed out that the central idea of "differences that make differences" represents the potential intervention (manipulation) in the cause-effect structure of the system [71]. We can also apply this argument to collective behaviour.…”

“…The aim of IIT 3.0, as we mentioned earlier, is to try to shift the paradigm from 'what the system does' to 'what the system is' [51,60,61,[69][70][71]. The former tries to analyse the system based on its external behaviour, whereas the latter tries to analyse the system according to its intrinsic causal structure.…”