"Indirection / A Model for Modules" and "Analytical chemistry; the science of many models"

[Donny Winston]

I loved the book. I noticed many happy parallels between the ideas of the "Indirection / A Model for Modules" section and a nine-page article by Żytkow and Lewenstam, "Analytical chemistry; the science of many models" (1990) <doi:10.1007/BF00323013>.

The article talks about composition of basic theories into models that are insufficient for application without considering structural statements about particular situations, i.e. assumptions of the environment. There are other nice mapping I think. The translation of modeled phenomena to a so-called model-diagram can map to the selection of data primitives and aggregate structures in programs. And the translation of model-diagram to so-called model-formalism of solvable equations can map to the the use of data in expressions and the composition of those expressions into tractable programs (whether declarative and thus equation-like or not) that are comparable with measurements of the modeled phenomenon (everything from automated regression tests to qualitative user interviews).

figure from the article

Here's one excerpt that highlights the importance of explicit description of the structure of the environment in which a set of basic theories are to be composed and reified as a model that reflects the environment:

> When we claim that scientific laws are deducible from basic laws, we use the verbalism of reductionist ideology, hiding the role of structural conditions and overdrawing the role of basic laws. Because both basic laws and structural conditions are required for the deductive inference of derived laws, can we claim that one is more important than the other? Knowledge of structure is equally as necessary as knowledge of basic laws. It is irreducible to basic laws, and as much effort has been required of mankind to reach the current knowledge of structure as it took to discover basic theories. In many domains scientists seek primarily knowledge about structure; for instance, the neural structure in the brain or the structure of Earth's interior. It is perhaps a truism for any scientist that we need to know the structure of a system if we want to apply the theory. But that simple truth is not properly acknowledged, and its consequences have little influence on the way in which scientists perceive their own work.

Interestingly, the lead author, Żytkow, co-authored a lot with Herbert Simon, associated with "satisfices" and GPS as mentioned in the book.

Great book, Zach. I'd be curious to know if you think the ideas of this article comport with your formulation, and if there's additional value here to be mapped to software modeling.

Best,

Donny

[Donny Winston]

Hmm, I realize that my pasting of a screenshot of the figure did not go through. Here is a link to me screenshot of the figure:

https://www.dropbox.com/s/9bhwsao2m6aokrc/zytkow-1990-modeling-flow.png?dl=0