Causal Informational Structural Realism

ABSTRACT

The debate between proponents and opponents of causal foundationalism has recently surfaced as a disparity between causal structuralism and causal anti-foundationalism in the structural realist camp. The paper outlines and dissolves the problem of disparity for (informational) structural realism. I follow John Collier (also Carl T. Bergstrom and Martin Rosvall) to specify causation in terms of the transmission of information. Unlike them, I built upon the reverse quantum data-processing inequality to show how this approach models causation as a symmetric process at the level of fundamental physics (but not special sciences). I show how this suggestion reduces the disparity about causation to a problem of application to diverse contexts.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1 For coming to the final version of this paper, I benefited from constructive comments of three anonymous reviewers of this journal, as well as Maria Panagiotatou from the editorial team. The debt is gratefully acknowledged.

2 Ladyman and Ross (2007, 220) specify logical depth as quantitative index of the execution time required to generate the model of the real patterns ‘by a near incompressible universal computer program, that is, one not itself computable as the output of a significantly more concise program’. It is a measure of the least computation time (or the least computational steps) required to compute an uncompressed string from its maximally compressed form (Collier 1999, 218 ff.)

3 Projectible relations are defined by saying that x → y is projectible if (i) there is a physically possible computing machine that could perform the projection x → y given some scale of resolution, and (ii) there is at least one other projection x → z that the computing machine can perform without changing its programme (Ladyman and Ross 2007, 224).

4 It is worth mentioning that casual structuralists usually support naturalism. Naturalism and CF do not need to be incompatible. This is because, it is not an established scientific fact that there is no causality in fundamental physics (in fact, metaphysical interpretation of laws of physics may bear on the question of existence/non-existence of causality at the level of fundamental physics (López 2019)). Moreover, causal realists’ naturalist tendency does not need to prevent them to retain parts of traditional metaphysics, for example by taking a Viking approach that allows them pillage and plunder what they find metaphysically precious (French 2014, section 3.2).

5 Calllender argues that even the Schrodinger equation is not originally time-reversal invariant, Eugene Wigner’s subtle modification notwithstanding (Callender 2000, 262; López 2019, 27 ff.).

6 A theory is called time-reversal invariant ‘iff for any history S_i, …; S_f allowed by the theory, the history S_f*, …; S_i* is also allowed by the theory, where S* is the time reverse of S, as determined by the time-reversal operator’ (Farr and Reutlinger 2013, 8). The definition does not indicate that two states that related by time reversal are the same states described in two different ways. Therefore, there is no contradiction between this conception of time-symmetry and directionality of causation.

7 Can be defined as,

$H\left(XY\right)=\sum _{xϵX}\sum _{yϵY}p\left(x,y\right)\mathrm{l}\mathrm{o}\mathrm{g}p\left(x,y\right)$

8 In their attempt to account for transmission of information by a minimal notion of causality, Lopez and Lombard too submit that ‘transmission of information is fundamentally an asymmetric phenomenon mounted on a causal structure’ (López and Lombardi 2018, 35), without specifically identifying asymmetry of transmission with time-asymmetry. Lopez and Lombardi also assume that communication is an asymmetric phenomena which involves causation. I take the reverse course to identify causation with transmission of information, and, as I explain in section 6, argue that causality could be symmetrical.

9 Spurious correlations or spurious causes could be defined in the following way: ‘An event A is a spurious cause of B if and only if A is a prima-facie cause of B, and there is a partition of events earlier than A such that the conditional probability of B, given A and any element of the partition, is the same as the conditional probability of B, given just the element of the partition’. (Suppes 1984, 50; Price 1991, 159).