As hinted at above, I don't think there really is a system in these tables (apart from some minor recurring sequences). Yet, there are different ways to go about committing them to memory. How would you memorize a table like the following?
I had an idea.
First of all, I tried to decide for myself which of the values were the most "basic" to me. I came to the conclusion that CONFIGURATION is a more basic category than EXTENSION (this is a subjective perception only), so I need to work outwards starting inside.
Next, there is a way to pretend that there is a system, or in other words, a compositionality present that really isn't there, but it has the effect of making the table look simpler than it is. Instead of taking the table as a single static block of information, one can treat the formation of the affix value like a path along a flowchart. Below I try to show how this works for an example Ca value.
The "→" symbol marks terminal nodes, that is, values for the Ca affix that you can plug into the Ca slot as is.
uniplex → l
associative → r
unbounded → k
proximal → g
In words, this flowchart path can be read as:
If the CONFIGURATON is uniplex, then the Ca affix has the value "l", unless its AFFILIATION is not consolidative; if it is associative, then the Ca affix has the value "r", unless the perspective is not monadic; if it is unbounded, then the Ca affix has the value "k", unless the EXTENSION is not delimitive; if it is proximal, the value of the Ca affix changes to "g"