Contents:
- A post-modern epiphany
- Reading Scaruffi: Gaining perspective on chapter 4
A Post-Modern Epiphany
One thing I’d like to note, real quick, is the analogy of Alice in Wonderland. It is possible to realize that all the rules and cultural institutions (e.g. library, bank, supermarket, school, army) are as sensical and straight-forward as are the Queen of Hearts and the rabbit in a waistcoat. This can be seen from multiple perspectives. It can be seen that nothing exists, and then It is, causing so that some things are true and some things are untrue; and from there, where the soul forgets it all except through its own self-discovery as it stares out into the stars losing eternity in their distance. An other (sic) possibility is that one inserts themselves into the enormous labyrinth of lies that mark the existence of our language in this world. And if the problem would be one of metaphysics of existence then it would be a subtle and esoteric conspiracy indeed. Instead, it is more obvious – the absurdity of it all; there is as much reason for respect as for the presence of Twiddledee and dum. (e.g. “I understand what they’re saying, but why are they there at all?”) In any case and whichever vista one peers down to reach this conclusion, any direct effect on praxis will come by the mould of the second of these two. Thus it is in the dreary everyday world that the greatest ludicrousness (cf. ludis: {Latin} a game, play) is to be found, providing a sense that the instructions are written in-jokes… how to describe this?… the feeling I get is as if I raise my head from a game of Snakes and Ladders in which I had become so engrossed as to forget the world beyond its scope and I then I look back at the board again – that feeling there that differentiates the two perceptions of the same game. I have another analogy for way of explanation, and this time it’s to describe my impression of the flow of realizations that go from seeing the world in its carnivalesque nature, and then that its shapes and boundaries are as real as any convention made up on the spot, that one can swim through them, like a fish, or a bird.
Reading Scaruffi: Gaining perspective on chapter 4
Although the focus of this chapter is clearly delineated in its title, i.e. cognition, I am struggling to see the forest for the trees. That is no surprise given the density of information in the book, but I hope to alleviate some of the pressure of immediacy by rephrasing some of the ideas here at a level of resolution, and with attention to form, that would best allow the clarity I seek.
(Numbering refers to my own counting of the subsections within the chapter).
[1] It is not too difficult to describe cognition as the focal point between the world taken as a sensation and the world as a target of the action.
[2] More difficult are attempts at explaining how that function (and note that what I described was a cognition qua function) comes to be mediated. After the preceding chapter 3, what is being sought here is something akin to an algorithm (e.g. instructions, or a category of instructions). An example of this is given by the classical paradigm of cognition that describes the bilateral interplay of knowledge and perception (i.e. the former is acquired by means of the latter, but the latter is processed on the basis of the former). It is noteworthy also that this paradigm employs the same concepts used to define cognition, and so suggests that in order to understand cognition it is that very relationship (i.e. between the mind as the receiver and the mind as a transceiver) which needs to be described or explained.
[3] It was implicit in the classical paradigm that the mind processes (and creates) representations (and also inferences). The is an important development on top of Descartes’ automaton-type being, who can act “as-if” it has intent; here, beings actually do have the minds they seem to have (i.e. the mind is not an illusory concept, useful but not real). This makes it clear that in describing cognition, we are treating it as a real phenomenon.
[4] Another paradigm is introduced: physical symbol systems. These are systems that process physical instantiations of symbols, and that the processing is also a process that acts in service of a goal. Here again is a suggested link to chapter 3: We can link what has been learned about machine intelligence (e.g. that symbolic logic can be incarnated as machine logic, and that the action of machine logic can form cybernetic systems) to the specifics of cognition [e.g. once Logic is accepted as a premise, it is possible to work/map towards cognition by adding a premises, including “goals”].
The paradigm of physical symbol systems is helpful for understanding the substance of cognition (i.e. cognition is made up of syntactical processes). But that remains a primitive step for understanding cognition, i.e. I would compare it with the explanation that cognition results from neurons and their activity. On top of [what I’ve called] the “substance” of cognition, is needed a model explaining how the units of cognition [presumably something like “information”] are processed and communicated within the cognitive being. This type of model was given basic form in the classical paradigm above. [5] A more sophisticated (but overrated and overused, according to Scaruffi) model is the creation of cognition by its production system.
Note the difference between the symbol system and this production system: (to repeat:) the substance upon which and by which cognition operates are symbols (according to this model), while the processes that are enacted upon these symbols can be described by the production system. Thus the two are mutually independent (i.e. at different levels of description) but also mutually compatible.
In skeletal form, a production system flows around a knowledge base; memory’s contents are operated upon by an inference engine, and the results funnelled into the knowledge base, while seen from the other side, the knowledge base is contained by the memory. Two forms of production systems are described in the text: ACT and SOAR.
Among the details of ACT’s (cognitive) architecture, is the relationship between declarative knowledge (i.e. “knowing that”) and procedural knowledge (i.e. “knowing how”). Note the parallel between this and both (i.) the dual term definition of cognition above, and (ii.) the dual pole description of cognition given by the classical paradigm. Here the focus is on explaining how action and behaviour can result: the production system not only interprets the declarative knowledge to produce action, but also transforms declarative knowledge into increasingly complex chunks of procedural knowledge (cf. chunking, but I would also note that this second point seems to be determining the difference between isolated action and sophisticated set-of-actions/behaviour).
SOAR is similar to ACT, and is also designed as a cognitive architecture that is best at coping with the information it receives from its environment (i.e. involving the achievement of goals). SOAR has three main components: (i.) universal weak method (i.e. knowledge determines methodology), (ii.) universal subgoaling (i.e. goals broken down into sub-goals as needed), and (iii.) chunking.
I note that ACT/SOAR meet the challenge raised by Scaruffi in the closing paragraph of chapter 3, where he wondered whether intelligence might be better sought as a means of problem identification, (rather than mere problem solving). Regardless as to what the creators of these architectures thought about intelligence and its definition, their models highlight the significance of goals for any paradigm (including cognition) that links input of information beyond one’s control to output whose range is as varied as one’s control. [To me this implies the fact that there can be no sense in that relationship without the introduction of direction, and that direction in the context of cognition is well described by “goals” – the source and nature of those goals is an entirely separate issue].
