More Precision

Let me go at this a bit more carefully. When we see familiar patterns of events we sometimes conclude that they are causally related, at least to some degree. Call such a pattern a ‘process’, at least here. Sometimes we can intervene in a process and convince ourselves more strongly that there are causes and effects within. Sometimes we can intervene in a process to detect causes and effects within.

The algorithm was a very early case of humans intending useful processes. The windmill includes a process. Institutions are processes whose elements are mostly humans.

We arrive thus at the category ‘process’. Some processes are deterministic, some random to a degree. We come thus to a 2nd level of category ‘deterministic’. Among deterministic processes there are calculations of π to which we shall soon return.

The digital computer can simulate many sorts of processes, certainly including other computers. With the commercial use of such things as the JVM (Java Virtual Machine) it has become practically necessary to dispense with distinctions between doing a calculation and the simulation thereof. Indeed one must often look very closely to make any such distinction. When we use Microsoft Office to produce a document, or OpenOffice to produce a like document, we conveniently elide the distinction that arises in the latter case where the real machine only runs the Java interpreter and does not really execute OpenOffice. With this elision we have created a category of process defined by the results it produces, regardless of intervening levels of simulation. I thus assert that a calculation of π and a simulation of a calculation of π are not to be distinguished; both processes produce digits of π.

I need a name for the category of process that ignores interpretive levels.