The Homunculus Theory of Cognition Always Appealed to Me

What with one thing and another, I have been spending a lot of my spare time in the last few months scrutinizing my mental processes.  The timing has been good, coinciding as it did with the meatier parts of my graduate school experience, in which the curriculum was equal parts basic engineering, information theory, and, relevantly, cognitive psychology.  So, at a time when I was learning to think about other people's mental experiences, I had a need to look to my own as well.  The crossover benefits are obvious.



At the same time, some of my decidedly uninformed philosophical opinions were coalescing, spurred partly by my conquest of A History of Western Philosophy, Bertrand's Russell's magnum opus.  One of the many luxuries it afforded me was the ability to identify myself as a materialist, in the sense that I believe that the observable material universe accessible to science is all that exists.  This is not a totally controversial position in this century, but it has some remarkable consequences; among them, that the myriad workings of the brain result only from physical processes, without the intervention of any metaphysical soul to clean up any questions of immortality or consciousness.  (I leave aside until never the question of free will.  It is an interesting question, to be sure, though possibly unanswerable, and in any case outside the scope of some dude's dusty blog.)  These processes, being governed only by physical laws that we mostly know, are therefore accessible to modelling by a sufficiently powerful computer; hence, any well-defined task or process we associate with cognitive, conscious beings can be performed by a computer.


As I am an avid (and now professional, a development new since my last post) programmer, it should be clear what interests me here, and thinking of the human brain as a highly sophisticated, powerful computer with a somewhat lacking QA process is very appealing.  In particular (and here, having summarized materialism, we come to what could be called the point), I enjoy discovering might would be called "undocumented features" if my brain had been developed by Google, say, or perhaps Canonical, or (saints preserve us) Research In Motion.  Due to a lamentable oversight, my brain did not come with any documentation at all, so all of its features, bugs, and so forth are necessarily undocumented; this does at least keep life interesting.


I discovered an undocumented feature this morning, the telling of which is the raison d'être of this entire post.  I was at the in-building coffee shop at work purchasing $3.87 in liquid nutriment, and I quite naturally paid with my debit card.  The following is what happened between myself and the cashier, from my perspective:

Cashier: Hands card back.

Dan: Receives card and starts to return it to wallet, in a well-practiced habitual motion. 

Cash Register: Makes ominous beeping sounds

C: "Cayeseetagin?" (Hurriedly) 

D: "I'm sorry?" Continues replacing card. 

C: "Cayeseetagin?" 

D: "Ah, of course."  Returns card. 

C: "It's just the machine."

What happened, a half-second after the cashier's second cryptic, perhaps Lovecraftian, utterance (no clearer than the first), is that I figured out that she had in fact been saying "Can I see it again?", referring, presumably, to my card.  I was not consciously struggling to parse her overlapping phonemes into any kind of sensible structure, and I certainly did not suddenly re-hear her utterance (in the mind's ear, if you like).  Rather, I simply, suddenly, knew what had been said a half-second earlier.


Clearly, something in my subconscious was keeping track of the fact that there was some outstanding sensory input that still needed to be processed, and it gave that a second look before disposing of it and attending to the next moment's information.  Rather than simply attempting to process information as it comes in, successfully or not, there is some process keeping track of success or failure that can give more processing to an important input that needs it.  The evolutionary benefits of such a system are so great that it seems obvious that it should be there, but it is the sort of obvious feature that I would never have thought of without having first experienced it.  Having experienced it, I took a moment to delight in my own inner workings, like a well-trained consumer cavorting with joy at the features of the newest, smartest smartphone.


I hope that my fascination with my self doesn't sound like narcissistic navel-gazing.  At least on a conscious level, my interest is joyfully academic rather than boastful.  I'm interested in how other people's brains work as well, but unfortunately mine is the only one I'm in a good position to observe at the moment.  It is most likely that it will come to nothing, being simply an omphalomantic exercise.  I dream, though, of the day that our understanding of human and non-human thought will be so thorough that we can create truly thinking beings in our own mental image, who will have the joy of discovering their own undocumented features.