Thread by @deontologistics

@elangelou A representation is a piece of information whose functional role is no longer understood merely in terms of success/failure but in terms of accuracy/inaccuracy. Representation is to psychology as function is to biology: analogies layered in making causal explanations/predictions.

@elangelou The analogy between function and practical reason is the foundation upon which we layer the analogy between representation and theoretical reason, thereby moving closer to describing the intentional structure of rational agency as such by bootstrapping our way out of analogy.

@elangelou This is what creates a distinction between roughly two types of representation (Kant: sensibility/understanding, Sellars: picturing/signification). The former is understood by analogy with the latter, even as the causal underpinnings of the latter are explained in terms of it.

@elangelou The aspect of the structure of theoretical reason on which representational analogies depend is that which Brandom makes explicit in his account of de re/de dicto vocabulary: the perspectival negotiation of discursive invariants that integrate potentially conflicting information.

@elangelou This gets reflected at the 'phenomenological' level of conscious intentionality in the functional structures negotiating perceptual invariants: capacities to track objects across time as sensory signals vary with positional perspective. These are really interactive invariants.

@elangelou The functional structure of conscious intentionality is cybernetic: there are loops between sensory input and behavioural output that track invariants in observation/manipulation of the environment (e.g., walking around an object so as to modulate one's own visual signal).

@elangelou The limitations of the representational analogy built on top of these functional structures are determined by the difference between abstract discursive interaction and concrete environmental interaction, e.g., between keeping track of anaphoric chains and embodied perspectives.

@elangelou There's a lot that could be said about this analogy between a perceptual object that can be viewed from multiple sensory perspectives and a discursive object that can be talked about from multiple inferential perspectives, but the crux what information they can integrate.

@elangelou An object in either sense is a locus of information integration, and should be understood in terms of how it structures potential incompatibilities between such information (e.g., both discursively and perceptually: the object 'x' cannot be both 'red' and 'green' all over).

@elangelou The really important thing about objects is that they can integrate different types of information (e.g., multi-modal sensory input). One can track the position of a moving car both visually and audibly, seamlessly integrating different sensory streams into spatial understanding.

@elangelou There are inherent limitations to our evolved capacities for acquiring and integrating sensory information. These are supplemented by 'conceptually formatted' representations that in virtue of their inferential articulation are extensible in principle: we can talk about 'quarks'.

@elangelou I could say more, but I think I've reached a nice equilibrium point: our ability to talk about our environment scientifically (e.g., 'threat vector') enables us to make an analogy with this ability in explaining the less flexible capacities that underpin it (e.g., [loud noise]).

@elangelou If anyone is interested, there's good reason to think that the natural structure of heterogeneous information integration, at least when we're talking about objects with 'fixed types' is given by sheaves (cf. sciencedirect.com/science/articl…).

@elangelou Another way to put this is that what Kant calls the understanding (the conceptual hierarchy qua type system) can be viewed as sheaves on a site (the forms of intuition qua base space). 'Space of possible experience' are Grothendieck topoi, whose internal logic is geometric.

@elangelou And, funnily enough, it turns out that Kant's transcendental logic (i.e., the logic of individuated objects that make judgments about them truth-apt) can be read as geometric logic: philpapers.org/rec/ACHAFO

@elangelou This doesn't get us all the way up to the faculty of reason, and how it allows us to change the conceptual frameworks upon which reference depends, but this is where dialectical logic emerges as a model of discursive invariants, and de re/de dicto talk is just the beginning.🖖

@elangelou CODA: If anyone is interested, here are some references to my work on this topic:

@elangelou 1. 'The Reformatting of Homo Sapiens' - which sketches a version of this methodology in terms of the evolution of animal and rational cognition: deontologistics.files.wordpress.com/2021/01/reform…

@elangelou 2. 'Is There a TV in My Head?: Content, Functional Mapping, and The Myth of the Given' - which digs into Sellars's account of perception in order to clarify the methodological issues involved in talking about the 'content' of perception: deontologistics.files.wordpress.com/2013/03/sellar…

@elangelou 3. 'Autonomy and Automation' - which takes the long way round to get to my account of general intelligence, but visits pretty much every step along the way:

@elangelou 4. 'Beyond Survival' - which provides the most elaborate version of my account of analogical bootstrapping:

@elangelou 5. The incomplete extended draft of 'Castalian Games' - which also takes a long path toward logic and dialectics, but contains a provisional, if dense account of how to map Sellars and Brandom's 'game of giving and asking for reasons' onto Girard's ludics: deontologistics.wordpress.com/wp-admin/uploa…

@elangelou 6. 'Copernicanism Without Correlationism' - which lays out my interpretation of Kant in more detail and provides a good background for everything said above: deontologistics.wordpress.com/2016/07/10/exp…

@elangelou Slides for all talks are available on my blog. I hope someone finds this useful. 🖖