Saturday, December 5, 2015

Walking in the Woods and Nested Frames of Reference

   I like to head out for a walk on Saturday and Sunday mornings. I get in the car and head west or north of south, as the case may be, and I enjoy my time in the car. I usually have a treat like a donut and coffee and I spend the time thinking about theoretical questions. Often lately I think about proto semantics and truisms but I also think about "best models". These theoretical thoughts stay with me in the woods for a while until the necessity of watching my footing, choosing a direction, and looking around overwhelm the theorizing and force me to contemplate the present moment. But in one case there is an overlap between the theoretical musings and the practical aspects of finding my way through the woods - a shared interest in how do I find my footing and choose my path?
   Put aside the 'larger' aspects of this question: why I am in this place? what direction I am headed? what is the best strategy for searching for rock piles [my basic motivation on weekend walks]? And consider the 'smaller' aspects of this question, as they help make the discussion as concrete as possible. These are the: where will I be in twenty feet? were will I take my next step? how shall my feet behave? I begin to get a sense of how these 'smaller' aspects of the question work.
   Take for example the last two: where is my next step and how shall my feet react? I already know where I am headed in the next twenty feet, so I pick my next step as part of keeping an even stride and avoiding difficult footing. If I see no obstructions my legs go on automatic. If I do and place my next step with some amount of consciousness, I usually defer to my feet to solve the problem of adapting to the specifics of what it encounters under foot. If the foot is "surprised" it calls back to the leg to get a different strategy for its next step. In the same way, if I cannot find a good next step it calls back to the process for choosing the next twenty feet and asks for a 'reset' to a different strategy. If I find my path blocked entirely - say by open water - then I may call back up to a higher process to change strategy.
   What these sorts of behaviors suggest to me is that there is a hierarchy of linked decision/action processes where each process determines a strategy and defers the tactics to a sub process. If the sub process encounters problems it calls back up to the parent and requests a new strategy. A process is a tactic for its parent and sets a strategy for its children. But it is of particular interest that this hierarchy is not simply a cognitive structure. There are specific perceptual aspects to each different level: I see twenty feet into the wood, I see beyond that to the larger "lay of the land", I see within it to where my feet will be placed. I perceive the footing with the same foot that is doing the stepping. The hierarchy is not just of cognition and of perception but of actuation as well.
   Each process has a strategy given to it and sets a tactic by selecting a sub process. Each process has an input perceptual aspect and an output actuation aspect.
   So that is it. This is not good writing but I hope it makes the point: handling the next twenty feet is done based on what I perceive in the next twenty feet and its actuation is to select a direction for the next step. Sub perception and sub actuations are deferred to and handled by the 'tactic' which is its own process. Thus no single process is responsible for the whole hierarchy; rather each part of the hierarchy takes responsibility for part of the perception and actuation needed for the task.

No comments:

Post a Comment