Section 4.5.1: Explanatory Gap Closed
5. Information-Theoretic Neutral-Structuralism: A Conjunction of Neutral Monism and Information-Theoretic Structural Realism
If we take the lessons from the previous four essays, then we are left with a rather difficult question which is the following: if we are to adopt the tools and concepts employed in a certain variety of information-theoretic explanation of consciousness – in this instance IIT – then what is the ontological and metaphysical picture we are left with? That is, should we want to resist falling into old traps of the battle between physicalism and anti-physicalism then we should look for a metaphysical framework that stays neutral in this fight – the consciousness wars’ equivalent of Switzerland stuck between the Allies and the Axis. What will follow are what I take to be two plausible candidates for such a project, one’s which come with their own set of troubles, but which I think are preferable to having the mind-body/hard problem looming as a consequence of one’s metaphysics. Both of these options will be ones that take on board the principle of informational realism, which is as follows:
Informational Realism (IR): “the view that the world is the totality of informational objects
dynamically interacting with each other (Floridi, 2004, p. 1) … instrumentally and predictively successful models (especially, but not only, those propounded by scientific theories) at a given LoA55 can be, in
the best circumstances, increasingly informative about the relations that obtain between the (possibly observable) informational objects that constitute the system under investigation (through the observable phenomena).” (Floridi, 2004, pp. 6–7)
Floridi borrows his notion of ‘objects’ in this context (when he speaks of ‘informational objects’) from object-oriented programming (OOP). The shift from thinking of representing data in an algorithmic sense, as something that takes an input, processes it, and gives some output as defined by the rules of the algorithm, to manipulating the objects as clusters of data; the difference being treating the data as the object to be manipulated rather than the logic procedures that govern it. Floridi uses the example
55 Floridi (2004) defines a level of abstraction (LoA) as “[a] LoA consists of a collection of observables. An observable is
an interpreted typed variable, that is, a variable with a well-defined possible set of values together with a statement of the properties of the system under consideration for which it stands. The target of a LoA is called a system. A system may be accessed and described at a range of LoAs and so can have a range of models.” A more detailed account of level of abstraction (LoA) can be found in Floridi (2011, chap. 3, 2008).
of a pawn in a chess game to elucidate this shift in thinking. There is a sense in which a pawn has a number of contingent features, shape, size, color, smoothness, etc., but those aren’t the properties one is concerned with when they play chess. Rather we care about the pawn’s strategic position on the board (what it can do and not do in relation to other pieces on the board), what possible moves it can make (it can move only one space at a time forward, attack diagonally, move two spaces if its first move, etc.). We treat pawns as informational objects in this sense according to Floridi, as he writes how OOP treats informational objects “…data structures (e.g. the pawn’s property of being white) and their behaviour (programming code, e.g. the pawn’s power to capture pieces only by moving diagonally forward) are packaged together as (informational) objects)” (Floridi, 2004, p. 5). Those informational objects are then hierarchically organized and inherit the characteristics of the level below, which are then deployed following some general rules for modelling at a particular level of abstraction. We can understand an informational object then as a cluster of data, the cluster of data having various well-defined characteristics, which are “differences de re, i.e. mind-independent points of lack of uniformity” (Floridi, 2004, p. 5). When I speak of an informational object/element/entity I am using it in the sense which Floridi does, though with one qualification: Floridi takes these objects in an ontological sense, and I take them as one useful way to talk about an informational object/entity/element.
I will however be arguing for a form of informational realism which is a conjunction of neutral monism and information-theoretic structural realism. I will argue in this essay that this gives us the following thesis:
Information-Theoretic Neutral-Structuralism: information is a mind-independent feature of reality,
not merely an abstraction from or as a result of physical and/or mental properties/processes/relationships, it is a distinct ontological category one which is prior to those properties, processes, relationships which we use the concepts ‘physical’ and ‘mental’ to describe. This nature is revealed through the structural relationships we discover in our natural investigation of the world, which point to patterns of informational structures. Put another way, following Russell, information is the common ancestor, neutral but prior.
This view would of course be in conflict with those who take information to be nothing over and above those physical representations (Landauer, 1996) which one might usually take as prior, more fundamental, etc., than information. The informational realism I adopt is also slightly weaker than Floridi’s as I am not committed to maintain a notion of information objects in the way Floridi
advocates, avoiding any kind of anthropocentric constraints on our ontology, as Floridi is committed with his notion of an informational object borrowed from OOP. Although I am not committed to informational objects in the way Floridi is, I nonetheless see the notion as useful in conceptualizing an informational object, though I land in favor of the view that such “objects” are merely points of lack of uniformity in the structure of the world.
The reader has most assuredly reached this point in the dissertation and felt that a firm definition of information is not only desired but necessary should one be basing their metaphysics on such a notion. I have, up until this point, resisted landing down firm on what such a notion of information is and how one should define it. This is largely due to the fact that I intended this dissertation to be something of an exploration of discovery into what such a notion of information might look like and how it might help us make sense of the natural world and the structure of reality, as well as, to shed light on consciousness and its place in nature. However, all expeditions must end, and this last essay is something of an expedition report and as a result I feel it best to lend my rough definition of information56 to help place down a mile marker for further investigation.
Information: A unique n-dimensional structure or shape in an abstract space of
possibilities which uniformly increases as novelty increases. The more unique or novel the shape of the structure/object/process the more information contained in that structure/object/process.
Here one may well see the heavy influence which IIT has had on this notion of information. This is largely done to accommodate lessons gleaned from IIT, in that when one unfolds the cause-effect structure of a system, one finds that the information contained in a system’s structural relations between its parts can in some instances be greater than those same parts taken individually. Also, accommodating Shannon’s original insight that information has to do with novelty or surprisal, that is the uncertainty that is reduced by a system being in a particular state or a particular message appearing given a string of letters. One may well see the apparent usefulness in this definition as well, should a system present with a high degree of organizational complexity then the information uniformly increases with such novel structure. With this now done we can set our sights on the main
56 I’m indebted to my external examiner Kelvin McQueen for his useful comments in needing to come down firmer on a
definition of information and for helping to organize my own thoughts on this matter, particularly the idea that information is a shape in a space of possibilities.
goal of this essay and that is the broader metaphysical/ontological implications of what has been discussed thus far.
This essay will be an investigation into what comes, given the context of coming up with an information-theoretic and natural account of phenomenal experience, if we ask the question “what is the nature of reality?” if we take information as our ontological starting point. If it can be meaningfully said that information can have an existence which is ontologically prior to those properties/processes/relationships which we normally speak of as physical or mental then I think we are in a position to present an interesting and novel metaphysical picture of the world as it concerns the problem of consciousness. This work will be a motivation of such an endeavor by examining two possible metaphysical pictures, the conjunction of which I think can accomplish such a task – those being (1) information-theoretic neutral monism (Sayre, 1976) and (2) information-theoretic structural realism (Ladyman et al., 2007). Each I think are interesting metaphysical alternatives to some of the options currently available to the philosopher of mind concerned with these topics and so I hope the conjunction of the two will offer a fresh take on these perpetually vexing issues of phenomenal experience and its place in the natural world.