[FRIAM] Formalizing the concept of design
∄ uǝʃƃ
gepropella at gmail.com
Sun Oct 28 06:53:18 EDT 2018
This description suffers from the same criticism I made before: you're assuming a *strict* hierarchy, where the higher order can only operate over whole components from the lower order. I.e. the gun's algorithm 1st chooses the type/medium of target (ballistic, air, water), then uses that type to select the specific tracking sub-algorithm.
And while this is mostly how it's done in artificial systems, I suspect biology does NOT use strict hierarchies. A higher order function can operate over a mixture of operands, some complex wholes in that higher order and some from the lower orders. E.g. if the gun's higher order selection is based not only on the 3 types (ballistic, air, water), but also on a lower order measure like *speed*, then it may well use he same sub-algorithm for both air and water. So, it takes both high order constructs and low order constructs as its operands.
You see your assumption of a strict hierarchy peeking through when you say sex is the only motive that is ESSENTIALLY social. What do you mean by "essentially"? Couldn't we say that *all* the behavior of all the social animals is, in part, social? ... including following others to the water hole? So, these functions would be mixed ... do not obey a strict hierarchy.
On 10/27/18 11:32 PM, Nick Thompson wrote:
> But the function that connects the two arrays will be different in the two kinds of gun because a surface target is capable of different sorts of motion from an aerial target.
> [...]
> So, the gun would display two levels of design, the lower level that relates trajectory to firing and the higher level that relates the lower level design to target type.
> [...]
> This conception of multiple hierarchical layers of design is a useful way to describe many of the phenomena that ethologists and socio-biologists are required to explain. …
--
∄ uǝʃƃ
More information about the Friam
mailing list