[Date Prev][Date Next][Thread Prev][Thread Next]   [Date Index] [Thread Index] [Author Index

Re: Modeling relations and semantics



A few comments:

Tim Gwinn wrote: My point is that the abstract mappings of the modeling
relation are what is a "free creation of the mind". If we *physically
realize* an abstract mapping by including use of a measuring device, it is
not that the measuring device per se is a "semantic element", but
snip
in a certain indirect sense, one can say that the measuring device (and
other associated physical aspects) are 'part' of the semantic nature of
encoding/decoding, but in my view it is only because those physical aspects
are part of the physical realization of the abstract mapping. I prefer to
say that the semantic aspect belongs to the abstract mapping, and not call
the physical devices used 'semantic'.
 
I would say that, where empirical/experimental/quantitative science is concerned, there are semantic elements all over the place. In this particular situation (creating a model of a natural system), we could say that semantic elements are natural by-products which come from the application of a measuring system, for the purposes of taking measurements of some aspect of the system. In other words, the activity of analysis, of deciding what to measure, how to measure (why to measure) and then, even more in the actual measuring and in putting the measurements together, etc.... There are semantic elements aplenty being generated in the entire process (encoding/decoding... model building). And none of these particular semantic elements is being generated from within the system we are measuring (encoding/decoding "how to build and test a model" is not entailed by the natural system). Neither is the process of building a model generated from within the model we end up building (encoding/decoding are not entailed by the formal system).
 
Every aspect of studying something (of developing ways to study it, ways to record what we think we are learning, ways of talking about the various aspects-- labels and names...) is an activity that is not entailed by the things we study nor is it entailed by the various tools and techniques we create in order to facilitate our endeavors.
 
So much of science takes place in the human mind, even with impirical science. The problem is that in our rather paranoid efforts to remove or eliminate the semantic pollution we add to any system we study, from perception and interpretation, all the way down to lab equipment and methods... we can end up erasing the natural semantic elements that the system DOES generate from within its own organization. Semantic elements are labeled "bad" and targeted for extermination. However, this is a mistake because in living systems, for example, semantic information represents the bulk of the information we need. Semantic can mean "relational". All relations are entirely context dependent for their values and in a highly integrated, interconnected system organization this means that all relations are also constantly interacting, which constantly changes context, and so on.
 
Context, itself, is a huge topic. There are innumerable values to context, because there are innumerable ways of defining it in any given situation. In other words, what we are labeling "context" in a scientific study of some system depends entirely on where we are observing from and what we are observing, etc.... depends entirely on contextual aspects of our observation.
 
The creation of any model we decide to build is going to have oodles of decisions about just exactly these sorts of things involved in the process, as well as all the other decisions we have to make about tools, measurements, techniques, modalities, etc.
 
Judith